From: Nicholas Clark Date: Sun, 18 Oct 2009 16:47:10 +0000 (+0100) Subject: Merge S_utf16_textfilter and S_utf16rev_textfilter(). X-Git-Url: http://git.shadowcat.co.uk/gitweb/gitweb.cgi?a=commitdiff_plain;h=aa6dbd607b0a3d8a49c43b63b63b73ed167692c3;p=p5sagit%2Fp5-mst-13.2.git Merge S_utf16_textfilter and S_utf16rev_textfilter(). Use IoLINES() on the filter's SV to determine which encoding is in use. --- diff --git a/embed.fnc b/embed.fnc index 7bbb91a..0fd0a41 100644 --- a/embed.fnc +++ b/embed.fnc @@ -1795,7 +1795,6 @@ sR |char* |skipspace |NN char *s sR |char* |swallow_bom |NN U8 *s #ifndef PERL_NO_UTF16_FILTER s |I32 |utf16_textfilter|int idx|NN SV *sv|int maxlen -s |I32 |utf16rev_textfilter|int idx|NN SV *sv|int maxlen #endif s |void |checkcomma |NN const char *s|NN const char *name \ |NN const char *what diff --git a/embed.h b/embed.h index 7e69fec..66c3194 100644 --- a/embed.h +++ b/embed.h @@ -1584,7 +1584,6 @@ #ifndef PERL_NO_UTF16_FILTER #ifdef PERL_CORE #define utf16_textfilter S_utf16_textfilter -#define utf16rev_textfilter S_utf16rev_textfilter #endif #endif #ifdef PERL_CORE @@ -3959,7 +3958,6 @@ #ifndef PERL_NO_UTF16_FILTER #ifdef PERL_CORE #define utf16_textfilter(a,b,c) S_utf16_textfilter(aTHX_ a,b,c) -#define utf16rev_textfilter(a,b,c) S_utf16rev_textfilter(aTHX_ a,b,c) #endif #endif #ifdef PERL_CORE diff --git a/proto.h b/proto.h index 69a6852..7d47e9b 100644 --- a/proto.h +++ b/proto.h @@ -5770,11 +5770,6 @@ STATIC I32 S_utf16_textfilter(pTHX_ int idx, SV *sv, int maxlen) #define PERL_ARGS_ASSERT_UTF16_TEXTFILTER \ assert(sv) -STATIC I32 S_utf16rev_textfilter(pTHX_ int idx, SV *sv, int maxlen) - __attribute__nonnull__(pTHX_2); -#define PERL_ARGS_ASSERT_UTF16REV_TEXTFILTER \ - assert(sv) - #endif STATIC void S_checkcomma(pTHX_ const char *s, const char *name, const char *what) __attribute__nonnull__(pTHX_1) diff --git a/toke.c b/toke.c index 8e47c9a..c7ec476 100644 --- a/toke.c +++ b/toke.c @@ -12708,7 +12708,7 @@ S_swallow_bom(pTHX_ U8 *s) U8 *news; I32 newlen; - filter_add(S_utf16rev_textfilter, NULL); + IoLINES(filter_add(S_utf16_textfilter, NULL)) = 1; Newx(news, (PL_bufend - (char*)s) * 3 / 2 + 1, U8); utf16_to_utf8_reversed(s, news, PL_bufend - (char*)s - 1, @@ -12810,8 +12810,10 @@ S_utf16_textfilter(pTHX_ int idx, SV *sv, int maxlen) dVAR; const STRLEN old = SvCUR(sv); const I32 count = FILTER_READ(idx+1, sv, maxlen); + const int reverse = IoLINES(sv); DEBUG_P(PerlIO_printf(Perl_debug_log, - "utf16_textfilter(%p): %d %d (%d)\n", + "utf16%s_textfilter(%p): %d %d (%d)\n", + reverse ? "rev" : "", FPTR2DPTR(void *, S_utf16_textfilter), idx, maxlen, (int) count)); if (count) { @@ -12819,36 +12821,22 @@ S_utf16_textfilter(pTHX_ int idx, SV *sv, int maxlen) I32 newlen; Newx(tmps, SvCUR(sv) * 3 / 2 + 1, U8); Copy(SvPVX_const(sv), tmps, old, char); - utf16_to_utf8((U8*)SvPVX_const(sv) + old, tmps + old, - SvCUR(sv) - old, &newlen); + if (reverse) { + /* You would expect this to be utf16_to_utf8_reversed() + It was, prior to 1de9afcdf18cf98bbdecaa782da93e907be6fe4e + Effectively, right now, UTF-16LE is being read in off-by-one + See RT #69678 */ + utf16_to_utf8((U8*)SvPVX_const(sv) + old, tmps + old, + SvCUR(sv) - old, &newlen); + } else { + utf16_to_utf8((U8*)SvPVX_const(sv) + old, tmps + old, + SvCUR(sv) - old, &newlen); + } sv_usepvn(sv, (char*)tmps, (STRLEN)newlen + old); } DEBUG_P({sv_dump(sv);}); return SvCUR(sv); } - -static I32 -S_utf16rev_textfilter(pTHX_ int idx, SV *sv, int maxlen) -{ - dVAR; - const STRLEN old = SvCUR(sv); - const I32 count = FILTER_READ(idx+1, sv, maxlen); - DEBUG_P(PerlIO_printf(Perl_debug_log, - "utf16rev_textfilter(%p): %d %d (%d)\n", - FPTR2DPTR(void *, utf16rev_textfilter), - idx, maxlen, (int) count)); - if (count) { - U8* tmps; - I32 newlen; - Newx(tmps, SvCUR(sv) * 3 / 2 + 1, U8); - Copy(SvPVX_const(sv), tmps, old, char); - utf16_to_utf8((U8*)SvPVX_const(sv) + old, tmps + old, - SvCUR(sv) - old, &newlen); - sv_usepvn(sv, (char*)tmps, (STRLEN)newlen + old); - } - DEBUG_P({ sv_dump(sv); }); - return count; -} #endif /*