X-Git-Url: https://perl5.git.perl.org/perl5.git/blobdiff_plain/043e41b864c9e8505919cb1347c4b340f7fd081b..2e38ced43368428d48ba9336fbc27d141ae9a761:/doop.c diff --git a/doop.c b/doop.c index dc525d3..892cd07 100644 --- a/doop.c +++ b/doop.c @@ -1,7 +1,7 @@ /* doop.c * - * Copyright (C) 1991, 1992, 1993, 1994, 1995, 1996, 1997, 1998, 1999, - * 2000, 2001, 2002, 2004, 2005, 2006, 2007, by Larry Wall and others + * Copyright (C) 1991, 1992, 1993, 1994, 1995, 1996, 1997, 1998, 1999, 2000, + * 2001, 2002, 2004, 2005, 2006, 2007, 2008, 2009 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. @@ -9,12 +9,14 @@ */ /* - * "'So that was the job I felt I had to do when I started,' thought Sam." + * 'So that was the job I felt I had to do when I started,' thought Sam. + * + * [p.934 of _The Lord of the Rings_, VI/iii: "Mount Doom"] */ /* This file contains some common functions needed to carry out certain - * ops. For example both pp_schomp() and pp_chomp() - scalar and array - * chomp operations - call the function do_chomp() found in this file. + * ops. For example, both pp_sprintf() and pp_prtf() call the function + * do_printf() found in this file. */ #include "EXTERN.h" @@ -28,13 +30,14 @@ STATIC I32 S_do_trans_simple(pTHX_ SV * const sv) { - dVAR; I32 matches = 0; STRLEN len; - U8 *s = (U8*)SvPV(sv,len); + U8 *s = (U8*)SvPV_nomg(sv,len); U8 * const send = s+len; - const short * const tbl = (short*)cPVOP->op_pv; + + PERL_ARGS_ASSERT_DO_TRANS_SIMPLE; + if (!tbl) Perl_croak(aTHX_ "panic: do_trans_simple line %d",__LINE__); @@ -95,13 +98,14 @@ S_do_trans_simple(pTHX_ SV * const sv) STATIC I32 S_do_trans_count(pTHX_ SV * const sv) { - dVAR; STRLEN len; - const U8 *s = (const U8*)SvPV_const(sv, len); + const U8 *s = (const U8*)SvPV_nomg_const(sv, len); const U8 * const send = s + len; I32 matches = 0; - const short * const tbl = (short*)cPVOP->op_pv; + + PERL_ARGS_ASSERT_DO_TRANS_COUNT; + if (!tbl) Perl_croak(aTHX_ "panic: do_trans_count line %d",__LINE__); @@ -131,13 +135,14 @@ S_do_trans_count(pTHX_ SV * const sv) STATIC I32 S_do_trans_complex(pTHX_ SV * const sv) { - dVAR; STRLEN len; - U8 *s = (U8*)SvPV(sv, len); + U8 *s = (U8*)SvPV_nomg(sv, len); U8 * const send = s+len; I32 matches = 0; - const short * const tbl = (short*)cPVOP->op_pv; + + PERL_ARGS_ASSERT_DO_TRANS_COMPLEX; + if (!tbl) Perl_croak(aTHX_ "panic: do_trans_complex line %d",__LINE__); @@ -195,10 +200,6 @@ S_do_trans_complex(pTHX_ SV * const sv) if (complement && !del) rlen = tbl[0x100]; -#ifdef MACOS_TRADITIONAL -#define comp CoMP /* "comp" is a keyword in some compilers ... */ -#endif - if (PL_op->op_private & OPpTRANS_SQUASH) { UV pch = 0xfeedface; while (s < send) { @@ -297,7 +298,6 @@ S_do_trans_complex(pTHX_ SV * const sv) STATIC I32 S_do_trans_simple_utf8(pTHX_ SV * const sv) { - dVAR; U8 *s; U8 *send; U8 *d; @@ -306,27 +306,28 @@ S_do_trans_simple_utf8(pTHX_ SV * const sv) I32 matches = 0; const I32 grows = PL_op->op_private & OPpTRANS_GROWS; STRLEN len; - SV* const rv = #ifdef USE_ITHREADS PAD_SVl(cPADOP->op_padix); #else - (SV*)cSVOP->op_sv; + MUTABLE_SV(cSVOP->op_sv); #endif - HV* const hv = (HV*)SvRV(rv); + HV* const hv = MUTABLE_HV(SvRV(rv)); SV* const * svp = hv_fetchs(hv, "NONE", FALSE); const UV none = svp ? SvUV(*svp) : 0x7fffffff; const UV extra = none + 1; UV final = 0; U8 hibit = 0; - s = (U8*)SvPV(sv, len); + PERL_ARGS_ASSERT_DO_TRANS_SIMPLE_UTF8; + + s = (U8*)SvPV_nomg(sv, len); if (!SvUTF8(sv)) { const U8 *t = s; const U8 * const e = s + len; while (t < e) { const U8 ch = *t++; - hibit = !NATIVE_IS_INVARIANT(ch); + hibit = !NATIVE_BYTE_IS_INVARIANT(ch); if (hibit) { s = bytes_to_utf8(s, &len); break; @@ -356,7 +357,7 @@ S_do_trans_simple_utf8(pTHX_ SV * const sv) if (uv < none) { s += UTF8SKIP(s); matches++; - d = uvuni_to_utf8(d, uv); + d = uvchr_to_utf8(d, uv); } else if (uv == none) { const int i = UTF8SKIP(s); @@ -367,7 +368,7 @@ S_do_trans_simple_utf8(pTHX_ SV * const sv) else if (uv == extra) { s += UTF8SKIP(s); matches++; - d = uvuni_to_utf8(d, final); + d = uvchr_to_utf8(d, final); } else s += UTF8SKIP(s); @@ -401,32 +402,32 @@ S_do_trans_simple_utf8(pTHX_ SV * const sv) STATIC I32 S_do_trans_count_utf8(pTHX_ SV * const sv) { - dVAR; const U8 *s; const U8 *start = NULL; const U8 *send; I32 matches = 0; STRLEN len; - SV* const rv = #ifdef USE_ITHREADS PAD_SVl(cPADOP->op_padix); #else - (SV*)cSVOP->op_sv; + MUTABLE_SV(cSVOP->op_sv); #endif - HV* const hv = (HV*)SvRV(rv); + HV* const hv = MUTABLE_HV(SvRV(rv)); SV* const * const svp = hv_fetchs(hv, "NONE", FALSE); const UV none = svp ? SvUV(*svp) : 0x7fffffff; const UV extra = none + 1; U8 hibit = 0; - s = (const U8*)SvPV_const(sv, len); + PERL_ARGS_ASSERT_DO_TRANS_COUNT_UTF8; + + s = (const U8*)SvPV_nomg_const(sv, len); if (!SvUTF8(sv)) { const U8 *t = s; const U8 * const e = s + len; while (t < e) { const U8 ch = *t++; - hibit = !NATIVE_IS_INVARIANT(ch); + hibit = !NATIVE_BYTE_IS_INVARIANT(ch); if (hibit) { start = s = bytes_to_utf8(s, &len); break; @@ -450,7 +451,6 @@ S_do_trans_count_utf8(pTHX_ SV * const sv) STATIC I32 S_do_trans_complex_utf8(pTHX_ SV * const sv) { - dVAR; U8 *start, *send; U8 *d; I32 matches = 0; @@ -461,9 +461,9 @@ S_do_trans_complex_utf8(pTHX_ SV * const sv) #ifdef USE_ITHREADS PAD_SVl(cPADOP->op_padix); #else - (SV*)cSVOP->op_sv; + MUTABLE_SV(cSVOP->op_sv); #endif - HV * const hv = (HV*)SvRV(rv); + HV * const hv = MUTABLE_HV(SvRV(rv)); SV * const *svp = hv_fetchs(hv, "NONE", FALSE); const UV none = svp ? SvUV(*svp) : 0x7fffffff; const UV extra = none + 1; @@ -472,14 +472,16 @@ S_do_trans_complex_utf8(pTHX_ SV * const sv) STRLEN len; U8 *dstart, *dend; U8 hibit = 0; + U8 *s = (U8*)SvPV_nomg(sv, len); + + PERL_ARGS_ASSERT_DO_TRANS_COMPLEX_UTF8; - U8 *s = (U8*)SvPV(sv, len); if (!SvUTF8(sv)) { const U8 *t = s; const U8 * const e = s + len; while (t < e) { const U8 ch = *t++; - hibit = !NATIVE_IS_INVARIANT(ch); + hibit = !NATIVE_BYTE_IS_INVARIANT(ch); if (hibit) { s = bytes_to_utf8(s, &len); break; @@ -524,7 +526,7 @@ S_do_trans_complex_utf8(pTHX_ SV * const sv) matches++; s += UTF8SKIP(s); if (uv != puv) { - d = uvuni_to_utf8(d, uv); + d = uvchr_to_utf8(d, uv); puv = uv; } continue; @@ -542,13 +544,13 @@ S_do_trans_complex_utf8(pTHX_ SV * const sv) if (havefinal) { s += UTF8SKIP(s); if (puv != final) { - d = uvuni_to_utf8(d, final); + d = uvchr_to_utf8(d, final); puv = final; } } else { STRLEN len; - uv = utf8n_to_uvuni(s, send - s, &len, UTF8_ALLOW_DEFAULT); + uv = utf8n_to_uvchr(s, send - s, &len, UTF8_ALLOW_DEFAULT); if (uv != puv) { Move(s, d, len, U8); d += len; @@ -577,7 +579,7 @@ S_do_trans_complex_utf8(pTHX_ SV * const sv) if (uv < none) { matches++; s += UTF8SKIP(s); - d = uvuni_to_utf8(d, uv); + d = uvchr_to_utf8(d, uv); continue; } else if (uv == none) { /* "none" is unmapped character */ @@ -590,7 +592,7 @@ S_do_trans_complex_utf8(pTHX_ SV * const sv) else if (uv == extra && !del) { matches++; s += UTF8SKIP(s); - d = uvuni_to_utf8(d, final); + d = uvchr_to_utf8(d, final); continue; } matches++; /* "none+1" is delete character */ @@ -616,63 +618,48 @@ S_do_trans_complex_utf8(pTHX_ SV * const sv) I32 Perl_do_trans(pTHX_ SV *sv) { - dVAR; STRLEN len; - const I32 hasutf = (PL_op->op_private & - (OPpTRANS_FROM_UTF|OPpTRANS_TO_UTF)); - - if (SvREADONLY(sv)) { - if (SvIsCOW(sv)) - sv_force_normal_flags(sv, 0); - if (SvREADONLY(sv) && !(PL_op->op_private & OPpTRANS_IDENTICAL)) - Perl_croak(aTHX_ PL_no_modify); + const I32 flags = PL_op->op_private; + const I32 hasutf = flags & (OPpTRANS_FROM_UTF | OPpTRANS_TO_UTF); + + PERL_ARGS_ASSERT_DO_TRANS; + + if (SvREADONLY(sv) && !(flags & OPpTRANS_IDENTICAL)) { + Perl_croak_no_modify(); } (void)SvPV_const(sv, len); if (!len) return 0; - if (!(PL_op->op_private & OPpTRANS_IDENTICAL)) { - if (!SvPOKp(sv)) - (void)SvPV_force(sv, len); + if (!(flags & OPpTRANS_IDENTICAL)) { + if (!SvPOKp(sv) || SvTHINKFIRST(sv)) + (void)SvPV_force_nomg(sv, len); (void)SvPOK_only_UTF8(sv); } DEBUG_t( Perl_deb(aTHX_ "2.TBL\n")); - switch (PL_op->op_private & ~hasutf & ( - OPpTRANS_FROM_UTF|OPpTRANS_TO_UTF|OPpTRANS_IDENTICAL| - OPpTRANS_SQUASH|OPpTRANS_DELETE|OPpTRANS_COMPLEMENT)) { - case 0: - if (hasutf) - return do_trans_simple_utf8(sv); - else - return do_trans_simple(sv); - - case OPpTRANS_IDENTICAL: - case OPpTRANS_IDENTICAL|OPpTRANS_COMPLEMENT: - if (hasutf) - return do_trans_count_utf8(sv); - else - return do_trans_count(sv); - - default: - if (hasutf) - return do_trans_complex_utf8(sv); - else - return do_trans_complex(sv); + /* If we use only OPpTRANS_IDENTICAL to bypass the READONLY check, + * we must also rely on it to choose the readonly strategy. + */ + if (flags & OPpTRANS_IDENTICAL) { + return hasutf ? do_trans_count_utf8(sv) : do_trans_count(sv); + } else if (flags & (OPpTRANS_SQUASH|OPpTRANS_DELETE|OPpTRANS_COMPLEMENT)) { + return hasutf ? do_trans_complex_utf8(sv) : do_trans_complex(sv); + } else { + return hasutf ? do_trans_simple_utf8(sv) : do_trans_simple(sv); } } void -Perl_do_join(pTHX_ register SV *sv, SV *delim, register SV **mark, register SV **sp) +Perl_do_join(pTHX_ SV *sv, SV *delim, SV **mark, SV **sp) { - dVAR; SV ** const oldmark = mark; - register I32 items = sp - mark; - register STRLEN len; + I32 items = sp - mark; + STRLEN len; STRLEN delimlen; + const char * const delims = SvPV_const(delim, delimlen); - (void) SvPV_const(delim, delimlen); /* stringify and get the delimlen */ - /* SvCUR assumes it's SvPOK() and woe betide you if it's not. */ + PERL_ARGS_ASSERT_DO_JOIN; mark++; len = (items > 0 ? (delimlen * (items - 1) ) : 0); @@ -693,11 +680,11 @@ Perl_do_join(pTHX_ register SV *sv, SV *delim, register SV **mark, register SV * ++mark; } - sv_setpvn(sv, "", 0); + sv_setpvs(sv, ""); /* sv_setpv retains old UTF8ness [perl #24846] */ SvUTF8_off(sv); - if (PL_tainting && SvMAGICAL(sv)) + if (TAINTING_get && SvMAGICAL(sv)) SvTAINTED_off(sv); if (items-- > 0) { @@ -707,14 +694,24 @@ Perl_do_join(pTHX_ register SV *sv, SV *delim, register SV **mark, register SV * } if (delimlen) { + const U32 delimflag = DO_UTF8(delim) ? SV_CATUTF8 : SV_CATBYTES; for (; items > 0; items--,mark++) { - sv_catsv(sv,delim); - sv_catsv(sv,*mark); + STRLEN len; + const char *s; + sv_catpvn_flags(sv,delims,delimlen,delimflag); + s = SvPV_const(*mark,len); + sv_catpvn_flags(sv,s,len, + DO_UTF8(*mark) ? SV_CATUTF8 : SV_CATBYTES); } } else { for (; items > 0; items--,mark++) - sv_catsv(sv,*mark); + { + STRLEN len; + const char *s = SvPV_const(*mark,len); + sv_catpvn_flags(sv,s,len, + DO_UTF8(*mark) ? SV_CATUTF8 : SV_CATBYTES); + } } SvSETMAGIC(sv); } @@ -722,11 +719,20 @@ Perl_do_join(pTHX_ register SV *sv, SV *delim, register SV **mark, register SV * void Perl_do_sprintf(pTHX_ SV *sv, I32 len, SV **sarg) { - dVAR; STRLEN patlen; const char * const pat = SvPV_const(*sarg, patlen); bool do_taint = FALSE; + PERL_ARGS_ASSERT_DO_SPRINTF; + + if (SvTAINTED(*sarg)) + TAINT_PROPER( + (PL_op && PL_op->op_type < OP_max) + ? (PL_op->op_type == OP_PRTF) + ? "printf" + : PL_op_name[PL_op->op_type] + : "(unknown)" + ); SvUTF8_off(sv); if (DO_UTF8(*sarg)) SvUTF8_on(sv); @@ -738,20 +744,31 @@ Perl_do_sprintf(pTHX_ SV *sv, I32 len, SV **sarg) /* currently converts input to bytes if possible, but doesn't sweat failure */ UV -Perl_do_vecget(pTHX_ SV *sv, I32 offset, I32 size) +Perl_do_vecget(pTHX_ SV *sv, SSize_t offset, int size) { - dVAR; STRLEN srclen, len, uoffset, bitoffs = 0; - const unsigned char *s = (const unsigned char *) SvPV_const(sv, srclen); + const I32 svpv_flags = ((PL_op->op_flags & OPf_MOD || LVRET) + ? SV_UNDEF_RETURNS_NULL : 0); + unsigned char *s = (unsigned char *) + SvPV_flags(sv, srclen, (svpv_flags|SV_GMAGIC)); UV retnum = 0; + if (!s) { + s = (unsigned char *)""; + } + + PERL_ARGS_ASSERT_DO_VECGET; + if (offset < 0) - return retnum; + return 0; if (size < 1 || (size & (size-1))) /* size < 1 or not a power of two */ Perl_croak(aTHX_ "Illegal number of bits in vec"); - if (SvUTF8(sv)) + if (SvUTF8(sv)) { (void) Perl_sv_utf8_downgrade(aTHX_ sv, TRUE); + /* PVX may have changed */ + s = (unsigned char *) SvPV_flags(sv, srclen, svpv_flags); + } if (size < 8) { bitoffs = ((offset%8)*size)%8; @@ -791,9 +808,8 @@ Perl_do_vecget(pTHX_ SV *sv, I32 offset, I32 size) } #ifdef UV_IS_QUAD else if (size == 64) { - if (ckWARN(WARN_PORTABLE)) - Perl_warner(aTHX_ packWARN(WARN_PORTABLE), - "Bit vector size > 32 non-portable"); + Perl_ck_warner(aTHX_ packWARN(WARN_PORTABLE), + "Bit vector size > 32 non-portable"); if (uoffset >= srclen) retnum = 0; else if (uoffset + 1 >= srclen) @@ -820,7 +836,7 @@ Perl_do_vecget(pTHX_ SV *sv, I32 offset, I32 size) ((UV) s[uoffset + 1] << 48) + ((UV) s[uoffset + 2] << 40) + ((UV) s[uoffset + 3] << 32) + - ( s[uoffset + 4] << 24); + ((UV) s[uoffset + 4] << 24); else if (uoffset + 6 >= srclen) retnum = ((UV) s[uoffset ] << 56) + @@ -837,7 +853,7 @@ Perl_do_vecget(pTHX_ SV *sv, I32 offset, I32 size) ((UV) s[uoffset + 3] << 32) + ((UV) s[uoffset + 4] << 24) + ((UV) s[uoffset + 5] << 16) + - ( s[uoffset + 6] << 8); + ((UV) s[uoffset + 6] << 8); } #endif } @@ -859,9 +875,8 @@ Perl_do_vecget(pTHX_ SV *sv, I32 offset, I32 size) s[uoffset + 3]; #ifdef UV_IS_QUAD else if (size == 64) { - if (ckWARN(WARN_PORTABLE)) - Perl_warner(aTHX_ packWARN(WARN_PORTABLE), - "Bit vector size > 32 non-portable"); + Perl_ck_warner(aTHX_ packWARN(WARN_PORTABLE), + "Bit vector size > 32 non-portable"); retnum = ((UV) s[uoffset ] << 56) + ((UV) s[uoffset + 1] << 48) + @@ -885,19 +900,21 @@ Perl_do_vecget(pTHX_ SV *sv, I32 offset, I32 size) void Perl_do_vecset(pTHX_ SV *sv) { - dVAR; - register I32 offset, bitoffs = 0; - register I32 size; - register unsigned char *s; - register UV lval; + SSize_t offset, bitoffs = 0; + int size; + unsigned char *s; + UV lval; I32 mask; STRLEN targlen; STRLEN len; SV * const targ = LvTARG(sv); + PERL_ARGS_ASSERT_DO_VECSET; + if (!targ) return; - s = (unsigned char*)SvPV_force(targ, targlen); + s = (unsigned char*)SvPV_force_flags(targ, targlen, + SV_GMAGIC | SV_UNDEF_RETURNS_NULL); if (SvUTF8(targ)) { /* This is handled by the SvPOK_only below... if (!Perl_sv_utf8_downgrade(aTHX_ targ, TRUE)) @@ -950,9 +967,8 @@ Perl_do_vecset(pTHX_ SV *sv) } #ifdef UV_IS_QUAD else if (size == 64) { - if (ckWARN(WARN_PORTABLE)) - Perl_warner(aTHX_ packWARN(WARN_PORTABLE), - "Bit vector size > 32 non-portable"); + Perl_ck_warner(aTHX_ packWARN(WARN_PORTABLE), + "Bit vector size > 32 non-portable"); s[offset ] = (U8)((lval >> 56) & 0xff); s[offset+1] = (U8)((lval >> 48) & 0xff); s[offset+2] = (U8)((lval >> 40) & 0xff); @@ -968,246 +984,41 @@ Perl_do_vecset(pTHX_ SV *sv) } void -Perl_do_chop(pTHX_ register SV *astr, register SV *sv) -{ - dVAR; - STRLEN len; - char *s; - - if (SvTYPE(sv) == SVt_PVAV) { - register I32 i; - AV* const av = (AV*)sv; - const I32 max = AvFILL(av); - - for (i = 0; i <= max; i++) { - sv = (SV*)av_fetch(av, i, FALSE); - if (sv && ((sv = *(SV**)sv), sv != &PL_sv_undef)) - do_chop(astr, sv); - } - return; - } - else if (SvTYPE(sv) == SVt_PVHV) { - HV* const hv = (HV*)sv; - HE* entry; - (void)hv_iterinit(hv); - while ((entry = hv_iternext(hv))) - do_chop(astr,hv_iterval(hv,entry)); - return; - } - else if (SvREADONLY(sv)) { - if (SvFAKE(sv)) { - /* SV is copy-on-write */ - sv_force_normal_flags(sv, 0); - } - if (SvREADONLY(sv)) - Perl_croak(aTHX_ PL_no_modify); - } - - if (PL_encoding && !SvUTF8(sv)) { - /* like in do_chomp(), utf8-ize the sv as a side-effect - * if we're using encoding. */ - sv_recode_to_utf8(sv, PL_encoding); - } - - s = SvPV(sv, len); - if (len && !SvPOK(sv)) - s = SvPV_force(sv, len); - if (DO_UTF8(sv)) { - if (s && len) { - char * const send = s + len; - char * const start = s; - s = send - 1; - while (s > start && UTF8_IS_CONTINUATION(*s)) - s--; - if (is_utf8_string((U8*)s, send - s)) { - sv_setpvn(astr, s, send - s); - *s = '\0'; - SvCUR_set(sv, s - start); - SvNIOK_off(sv); - SvUTF8_on(astr); - } - } - else - sv_setpvn(astr, "", 0); - } - else if (s && len) { - s += --len; - sv_setpvn(astr, s, 1); - *s = '\0'; - SvCUR_set(sv, len); - SvUTF8_off(sv); - SvNIOK_off(sv); - } - else - sv_setpvn(astr, "", 0); - SvSETMAGIC(sv); -} - -I32 -Perl_do_chomp(pTHX_ register SV *sv) -{ - dVAR; - register I32 count; - STRLEN len; - char *s; - char *temp_buffer = NULL; - SV* svrecode = NULL; - - if (RsSNARF(PL_rs)) - return 0; - if (RsRECORD(PL_rs)) - return 0; - count = 0; - if (SvTYPE(sv) == SVt_PVAV) { - register I32 i; - AV* const av = (AV*)sv; - const I32 max = AvFILL(av); - - for (i = 0; i <= max; i++) { - sv = (SV*)av_fetch(av, i, FALSE); - if (sv && ((sv = *(SV**)sv), sv != &PL_sv_undef)) - count += do_chomp(sv); - } - return count; - } - else if (SvTYPE(sv) == SVt_PVHV) { - HV* const hv = (HV*)sv; - HE* entry; - (void)hv_iterinit(hv); - while ((entry = hv_iternext(hv))) - count += do_chomp(hv_iterval(hv,entry)); - return count; - } - else if (SvREADONLY(sv)) { - if (SvFAKE(sv)) { - /* SV is copy-on-write */ - sv_force_normal_flags(sv, 0); - } - if (SvREADONLY(sv)) - Perl_croak(aTHX_ PL_no_modify); - } - - if (PL_encoding) { - if (!SvUTF8(sv)) { - /* XXX, here sv is utf8-ized as a side-effect! - If encoding.pm is used properly, almost string-generating - operations, including literal strings, chr(), input data, etc. - should have been utf8-ized already, right? - */ - sv_recode_to_utf8(sv, PL_encoding); - } - } - - s = SvPV(sv, len); - if (s && len) { - s += --len; - if (RsPARA(PL_rs)) { - if (*s != '\n') - goto nope; - ++count; - while (len && s[-1] == '\n') { - --len; - --s; - ++count; - } - } - else { - STRLEN rslen, rs_charlen; - const char *rsptr = SvPV_const(PL_rs, rslen); - - rs_charlen = SvUTF8(PL_rs) - ? sv_len_utf8(PL_rs) - : rslen; - - if (SvUTF8(PL_rs) != SvUTF8(sv)) { - /* Assumption is that rs is shorter than the scalar. */ - if (SvUTF8(PL_rs)) { - /* RS is utf8, scalar is 8 bit. */ - bool is_utf8 = TRUE; - temp_buffer = (char*)bytes_from_utf8((U8*)rsptr, - &rslen, &is_utf8); - if (is_utf8) { - /* Cannot downgrade, therefore cannot possibly match - */ - assert (temp_buffer == rsptr); - temp_buffer = NULL; - goto nope; - } - rsptr = temp_buffer; - } - else if (PL_encoding) { - /* RS is 8 bit, encoding.pm is used. - * Do not recode PL_rs as a side-effect. */ - svrecode = newSVpvn(rsptr, rslen); - sv_recode_to_utf8(svrecode, PL_encoding); - rsptr = SvPV_const(svrecode, rslen); - rs_charlen = sv_len_utf8(svrecode); - } - else { - /* RS is 8 bit, scalar is utf8. */ - temp_buffer = (char*)bytes_to_utf8((U8*)rsptr, &rslen); - rsptr = temp_buffer; - } - } - if (rslen == 1) { - if (*s != *rsptr) - goto nope; - ++count; - } - else { - if (len < rslen - 1) - goto nope; - len -= rslen - 1; - s -= rslen - 1; - if (memNE(s, rsptr, rslen)) - goto nope; - count += rs_charlen; - } - } - s = SvPV_force_nolen(sv); - SvCUR_set(sv, len); - *SvEND(sv) = '\0'; - SvNIOK_off(sv); - SvSETMAGIC(sv); - } - nope: - - if (svrecode) - SvREFCNT_dec(svrecode); - - Safefree(temp_buffer); - return count; -} - -void Perl_do_vop(pTHX_ I32 optype, SV *sv, SV *left, SV *right) { - dVAR; #ifdef LIBERAL - register long *dl; - register long *ll; - register long *rl; + long *dl; + long *ll; + long *rl; #endif - register char *dc; + char *dc; STRLEN leftlen; STRLEN rightlen; - register const char *lc; - register const char *rc; - register STRLEN len; + const char *lc; + const char *rc; + STRLEN len; STRLEN lensave; const char *lsave; const char *rsave; bool left_utf; bool right_utf; + bool do_warn_above_ff = ckWARN_d(WARN_DEPRECATED); STRLEN needlen = 0; + PERL_ARGS_ASSERT_DO_VOP; - if (sv != left || (optype != OP_BIT_AND && !SvOK(sv) && !SvGMAGICAL(sv))) - sv_setpvn(sv, "", 0); /* avoid undef warning on |= and ^= */ - lsave = lc = SvPV_nomg_const(left, leftlen); + if (sv != left || (optype != OP_BIT_AND && !SvOK(sv))) + sv_setpvs(sv, ""); /* avoid undef warning on |= and ^= */ + if (sv == left) { + lsave = lc = SvPV_force_nomg(left, leftlen); + } + else { + lsave = lc = SvPV_nomg_const(left, leftlen); + SvPV_force_nomg_nolen(sv); + } rsave = rc = SvPV_nomg_const(right, rightlen); - /* This need to come after SvPV to ensure that string overloading has + /* This needs to come after SvPV to ensure that string overloading has fired off. */ left_utf = DO_UTF8(left); @@ -1217,13 +1028,13 @@ Perl_do_vop(pTHX_ I32 optype, SV *sv, SV *left, SV *right) /* Avoid triggering overloading again by using temporaries. Maybe there should be a variant of sv_utf8_upgrade that takes pvn */ - right = sv_2mortal(newSVpvn(rsave, rightlen)); + right = newSVpvn_flags(rsave, rightlen, SVs_TEMP); sv_utf8_upgrade(right); rsave = rc = SvPV_nomg_const(right, rightlen); right_utf = TRUE; } else if (!left_utf && right_utf) { - left = sv_2mortal(newSVpvn(lsave, leftlen)); + left = newSVpvn_flags(lsave, leftlen, SVs_TEMP); sv_utf8_upgrade(left); lsave = lc = SvPV_nomg_const(left, leftlen); left_utf = TRUE; @@ -1272,6 +1083,12 @@ Perl_do_vop(pTHX_ I32 optype, SV *sv, SV *left, SV *right) rulen -= ulen; duc = luc & ruc; dc = (char*)uvchr_to_utf8((U8*)dc, duc); + if (do_warn_above_ff && (luc > 0xff || ruc > 0xff)) { + Perl_warner(aTHX_ packWARN(WARN_DEPRECATED), + deprecated_above_ff_msg, PL_op_desc[optype]); + /* Warn only once per operation */ + do_warn_above_ff = FALSE; + } } if (sv == left || sv == right) (void)sv_usepvn(sv, dcorig, needlen); @@ -1287,6 +1104,11 @@ Perl_do_vop(pTHX_ I32 optype, SV *sv, SV *left, SV *right) rulen -= ulen; duc = luc ^ ruc; dc = (char*)uvchr_to_utf8((U8*)dc, duc); + if (do_warn_above_ff && (luc > 0xff || ruc > 0xff)) { + Perl_warner(aTHX_ packWARN(WARN_DEPRECATED), + deprecated_above_ff_msg, PL_op_desc[optype]); + do_warn_above_ff = FALSE; + } } goto mop_up_utf; case OP_BIT_OR: @@ -1299,6 +1121,11 @@ Perl_do_vop(pTHX_ I32 optype, SV *sv, SV *left, SV *right) rulen -= ulen; duc = luc | ruc; dc = (char*)uvchr_to_utf8((U8*)dc, duc); + if (do_warn_above_ff && (luc > 0xff || ruc > 0xff)) { + Perl_warner(aTHX_ packWARN(WARN_DEPRECATED), + deprecated_above_ff_msg, PL_op_desc[optype]); + do_warn_above_ff = FALSE; + } } mop_up_utf: if (rulen) @@ -1306,12 +1133,12 @@ Perl_do_vop(pTHX_ I32 optype, SV *sv, SV *left, SV *right) else if (lulen) dcsave = savepvn(lc, lulen); if (sv == left || sv == right) - (void)sv_usepvn(sv, dcorig, needlen); /* Uses Renew(). */ + (void)sv_usepvn(sv, dcorig, needlen); /* uses Renew(); defaults to nomg */ SvCUR_set(sv, dc - dcorig); if (rulen) - sv_catpvn(sv, dcsave, rulen); + sv_catpvn_nomg(sv, dcsave, rulen); else if (lulen) - sv_catpvn(sv, dcsave, lulen); + sv_catpvn_nomg(sv, dcsave, lulen); else *SvEND(sv) = '\0'; Safefree(dcsave); @@ -1389,109 +1216,85 @@ Perl_do_vop(pTHX_ I32 optype, SV *sv, SV *left, SV *right) mop_up: len = lensave; if (rightlen > len) - sv_catpvn(sv, rsave + len, rightlen - len); + sv_catpvn_nomg(sv, rsave + len, rightlen - len); else if (leftlen > (STRLEN)len) - sv_catpvn(sv, lsave + len, leftlen - len); + sv_catpvn_nomg(sv, lsave + len, leftlen - len); else *SvEND(sv) = '\0'; break; } } -finish: + finish: SvTAINT(sv); } + +/* used for: pp_keys(), pp_values() */ + OP * Perl_do_kv(pTHX) { - dVAR; dSP; - HV * const hv = (HV*)POPs; - HV *keys; - register HE *entry; - const I32 gimme = GIMME_V; + HV * const keys = MUTABLE_HV(POPs); + HE *entry; + SSize_t extend_size; + const U8 gimme = GIMME_V; const I32 dokv = (PL_op->op_type == OP_RV2HV || PL_op->op_type == OP_PADHV); + /* op_type is OP_RKEYS/OP_RVALUES if pp_rkeys delegated to here */ const I32 dokeys = dokv || (PL_op->op_type == OP_KEYS); const I32 dovalues = dokv || (PL_op->op_type == OP_VALUES); - if (!hv) { - if (PL_op->op_flags & OPf_MOD || LVRET) { /* lvalue */ - dTARGET; /* make sure to clear its target here */ - if (SvTYPE(TARG) == SVt_PVLV) - LvTARG(TARG) = NULL; - PUSHs(TARG); - } - RETURN; - } - - keys = hv; (void)hv_iterinit(keys); /* always reset iterator regardless */ if (gimme == G_VOID) RETURN; if (gimme == G_SCALAR) { - IV i; - dTARGET; - if (PL_op->op_flags & OPf_MOD || LVRET) { /* lvalue */ - if (SvTYPE(TARG) < SVt_PVLV) { - sv_upgrade(TARG, SVt_PVLV); - sv_magic(TARG, NULL, PERL_MAGIC_nkeys, NULL, 0); - } - LvTYPE(TARG) = 'k'; - if (LvTARG(TARG) != (SV*)keys) { - if (LvTARG(TARG)) - SvREFCNT_dec(LvTARG(TARG)); - LvTARG(TARG) = SvREFCNT_inc_simple(keys); - } - PUSHs(TARG); - RETURN; - } - - if (! SvTIED_mg((SV*)keys, PERL_MAGIC_tied) ) - { - i = HvKEYS(keys); + SV * const ret = sv_2mortal(newSV_type(SVt_PVLV)); /* Not TARG RT#67838 */ + sv_magic(ret, NULL, PERL_MAGIC_nkeys, NULL, 0); + LvTYPE(ret) = 'k'; + LvTARG(ret) = SvREFCNT_inc_simple(keys); + PUSHs(ret); } else { - i = 0; - while (hv_iternext(keys)) i++; + IV i; + dTARGET; + + if (! SvTIED_mg((const SV *)keys, PERL_MAGIC_tied) ) { + i = HvUSEDKEYS(keys); + } + else { + i = 0; + while (hv_iternext(keys)) i++; + } + PUSHi( i ); } - PUSHi( i ); RETURN; } - EXTEND(SP, HvKEYS(keys) * (dokeys + dovalues)); + /* 2*HvUSEDKEYS() should never be big enough to truncate or wrap */ + assert(HvUSEDKEYS(keys) <= (SSize_t_MAX >> 1)); + extend_size = (SSize_t)HvUSEDKEYS(keys) * (dokeys + dovalues); + EXTEND(SP, extend_size); - PUTBACK; /* hv_iternext and hv_iterval might clobber stack_sp */ while ((entry = hv_iternext(keys))) { - SPAGAIN; if (dokeys) { SV* const sv = hv_iterkeysv(entry); - XPUSHs(sv); /* won't clobber stack_sp */ + XPUSHs(sv); } if (dovalues) { - SV *tmpstr; - PUTBACK; - tmpstr = hv_iterval(hv,entry); + SV *tmpstr = hv_iterval(keys,entry); DEBUG_H(Perl_sv_setpvf(aTHX_ tmpstr, "%lu%%%d=%lu", (unsigned long)HeHASH(entry), (int)HvMAX(keys)+1, (unsigned long)(HeHASH(entry) & HvMAX(keys)))); - SPAGAIN; XPUSHs(tmpstr); } - PUTBACK; } - return NORMAL; + RETURN; } /* - * Local variables: - * c-indentation-style: bsd - * c-basic-offset: 4 - * indent-tabs-mode: t - * End: - * - * ex: set ts=8 sts=4 sw=4 noet: + * ex: set ts=8 sts=4 sw=4 et: */