From: Hugo van der Sanden Date: Tue, 5 Nov 2002 13:30:00 +0000 (+0000) Subject: Re: [perl #18107] lc(), uc() and ucfirst() broken inside utf8 regex X-Git-Tag: accepted/trunk/20130322.191538~25163 X-Git-Url: http://review.tizen.org/git/?a=commitdiff_plain;h=d104a74c650f174423937daa379e5578a56dc3bb;p=platform%2Fupstream%2Fperl.git Re: [perl #18107] lc(), uc() and ucfirst() broken inside utf8 regex Message-Id: <200211051330.gA5DU0p00718@crypt.compulink.co.uk> This is half of the fix for #18107. The swash_init() problems described in <20021105143646.M18101@lustre.dyn.wiw.org> remain, but can now be seen only with UTF-8 test cases. p4raw-id: //depot/perl@18109 --- diff --git a/pp.c b/pp.c index 4c05aab..1a1dcea 100644 --- a/pp.c +++ b/pp.c @@ -3323,7 +3323,8 @@ PP(pp_ucfirst) register U8 *s; STRLEN slen; - if (DO_UTF8(sv) && (s = (U8*)SvPV(sv, slen)) && slen && UTF8_IS_START(*s)) { + SvGETMAGIC(sv); + if (DO_UTF8(sv) && (s = (U8*)SvPV_nomg(sv, slen)) && slen && UTF8_IS_START(*s)) { U8 tmpbuf[UTF8_MAXLEN_UCLC+1]; STRLEN ulen; STRLEN tculen; @@ -3340,7 +3341,7 @@ PP(pp_ucfirst) SETs(TARG); } else { - s = (U8*)SvPV_force(sv, slen); + s = (U8*)SvPV_force_nomg(sv, slen); Copy(tmpbuf, s, tculen, U8); } } @@ -3348,11 +3349,11 @@ PP(pp_ucfirst) if (!SvPADTMP(sv) || SvREADONLY(sv)) { dTARGET; SvUTF8_off(TARG); /* decontaminate */ - sv_setsv(TARG, sv); + sv_setsv_nomg(TARG, sv); sv = TARG; SETs(sv); } - s = (U8*)SvPV_force(sv, slen); + s = (U8*)SvPV_force_nomg(sv, slen); if (*s) { if (IN_LOCALE_RUNTIME) { TAINT; @@ -3363,8 +3364,7 @@ PP(pp_ucfirst) *s = toUPPER(*s); } } - if (SvSMAGICAL(sv)) - mg_set(sv); + SvSETMAGIC(sv); RETURN; } @@ -3375,7 +3375,8 @@ PP(pp_lcfirst) register U8 *s; STRLEN slen; - if (DO_UTF8(sv) && (s = (U8*)SvPV(sv, slen)) && slen && UTF8_IS_START(*s)) { + SvGETMAGIC(sv); + if (DO_UTF8(sv) && (s = (U8*)SvPV_nomg(sv, slen)) && slen && UTF8_IS_START(*s)) { STRLEN ulen; U8 tmpbuf[UTF8_MAXLEN_UCLC+1]; U8 *tend; @@ -3393,7 +3394,7 @@ PP(pp_lcfirst) SETs(TARG); } else { - s = (U8*)SvPV_force(sv, slen); + s = (U8*)SvPV_force_nomg(sv, slen); Copy(tmpbuf, s, ulen, U8); } } @@ -3401,11 +3402,11 @@ PP(pp_lcfirst) if (!SvPADTMP(sv) || SvREADONLY(sv)) { dTARGET; SvUTF8_off(TARG); /* decontaminate */ - sv_setsv(TARG, sv); + sv_setsv_nomg(TARG, sv); sv = TARG; SETs(sv); } - s = (U8*)SvPV_force(sv, slen); + s = (U8*)SvPV_force_nomg(sv, slen); if (*s) { if (IN_LOCALE_RUNTIME) { TAINT; @@ -3416,8 +3417,7 @@ PP(pp_lcfirst) *s = toLOWER(*s); } } - if (SvSMAGICAL(sv)) - mg_set(sv); + SvSETMAGIC(sv); RETURN; } @@ -3428,6 +3428,7 @@ PP(pp_uc) register U8 *s; STRLEN len; + SvGETMAGIC(sv); if (DO_UTF8(sv)) { dTARGET; STRLEN ulen; @@ -3435,7 +3436,7 @@ PP(pp_uc) U8 *send; U8 tmpbuf[UTF8_MAXLEN_UCLC+1]; - s = (U8*)SvPV(sv,len); + s = (U8*)SvPV_nomg(sv,len); if (!len) { SvUTF8_off(TARG); /* decontaminate */ sv_setpvn(TARG, "", 0); @@ -3465,11 +3466,11 @@ PP(pp_uc) if (!SvPADTMP(sv) || SvREADONLY(sv)) { dTARGET; SvUTF8_off(TARG); /* decontaminate */ - sv_setsv(TARG, sv); + sv_setsv_nomg(TARG, sv); sv = TARG; SETs(sv); } - s = (U8*)SvPV_force(sv, len); + s = (U8*)SvPV_force_nomg(sv, len); if (len) { register U8 *send = s + len; @@ -3485,8 +3486,7 @@ PP(pp_uc) } } } - if (SvSMAGICAL(sv)) - mg_set(sv); + SvSETMAGIC(sv); RETURN; } @@ -3497,6 +3497,7 @@ PP(pp_lc) register U8 *s; STRLEN len; + SvGETMAGIC(sv); if (DO_UTF8(sv)) { dTARGET; STRLEN ulen; @@ -3504,7 +3505,7 @@ PP(pp_lc) U8 *send; U8 tmpbuf[UTF8_MAXLEN_UCLC+1]; - s = (U8*)SvPV(sv,len); + s = (U8*)SvPV_nomg(sv,len); if (!len) { SvUTF8_off(TARG); /* decontaminate */ sv_setpvn(TARG, "", 0); @@ -3551,12 +3552,12 @@ PP(pp_lc) if (!SvPADTMP(sv) || SvREADONLY(sv)) { dTARGET; SvUTF8_off(TARG); /* decontaminate */ - sv_setsv(TARG, sv); + sv_setsv_nomg(TARG, sv); sv = TARG; SETs(sv); } - s = (U8*)SvPV_force(sv, len); + s = (U8*)SvPV_force_nomg(sv, len); if (len) { register U8 *send = s + len; @@ -3572,8 +3573,7 @@ PP(pp_lc) } } } - if (SvSMAGICAL(sv)) - mg_set(sv); + SvSETMAGIC(sv); RETURN; }