1berry/ruby
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

* transcode.c (str_transcode0): disable newline conversion for ASCII

Browse Source 
incompatible encoding.



git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@18853 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
This commit is contained in:
akr 2008-08-25 16:27:23 +00:00
parent 1a805010ed
commit 873987f2e0
2 changed files with 27 additions and 7 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

5
ChangeLog
View File

@ -1,3 +1,8 @@
Tue Aug 26 01:26:31 2008 Tanaka Akira <akr@fsij.org>
* transcode.c (str_transcode0): disable newline conversion for ASCII
incompatible encoding.
Tue Aug 26 00:55:46 2008 Tanaka Akira <akr@fsij.org> Tue Aug 26 00:55:46 2008 Tanaka Akira <akr@fsij.org>
* transcode.c (str_transcode0): don't short cut for newline conversion. * transcode.c (str_transcode0): don't short cut for newline conversion.

29
transcode.c
View File

@ -1861,7 +1861,7 @@ str_transcode_enc_args(VALUE str, VALUE arg1, VALUE arg2,
} }
static int static int
str_transcode0(int argc, VALUE *argv, VALUE *self, rb_econv_option_t *ecopts) str_transcode0(int argc, VALUE *argv, VALUE *self, rb_econv_option_t *ecopts_arg)
{ {
VALUE dest; VALUE dest;
VALUE str = *self; VALUE str = *self;
@ -1871,6 +1871,7 @@ str_transcode0(int argc, VALUE *argv, VALUE *self, rb_econv_option_t *ecopts)
rb_encoding *from_enc, *to_enc; rb_encoding *from_enc, *to_enc;
const char *from_e, *to_e; const char *from_e, *to_e;
int to_encidx; int to_encidx;
rb_econv_option_t ecopts;
if (argc < 1 || argc > 2) { if (argc < 1 || argc > 2) {
rb_raise(rb_eArgError, "wrong number of arguments (%d for 1..2)", argc); rb_raise(rb_eArgError, "wrong number of arguments (%d for 1..2)", argc);
@ -1878,9 +1879,25 @@ str_transcode0(int argc, VALUE *argv, VALUE *self, rb_econv_option_t *ecopts)
to_encidx = str_transcode_enc_args(str, argv[0], argc==1 ? Qnil : argv[1], &from_e, &from_enc, &to_e, &to_enc); to_encidx = str_transcode_enc_args(str, argv[0], argc==1 ? Qnil : argv[1], &from_e, &from_enc, &to_e, &to_enc);
if ((ecopts->flags & (ECONV_UNIVERSAL_NEWLINE_DECODER| if (ecopts_arg)
ECONV_CRLF_NEWLINE_ENCODER| ecopts = *ecopts_arg;
ECONV_CR_NEWLINE_ENCODER)) == 0) { else
rb_econv_opts(Qnil, &ecopts);
/* disable newline conversion for ascii incompatible encoding.
* xxx: convert newline in ascii-compatible encoding?
* ex. UTF-16BE -> UTF-8 -> newline conversion -> UTF-8 -> UTF-16BE.
*/
if (!from_enc || !rb_enc_asciicompat(from_enc)) {
ecopts.flags &= ~(ECONV_CRLF_NEWLINE_ENCODER|ECONV_CR_NEWLINE_ENCODER);
}
if (!to_enc || !rb_enc_asciicompat(to_enc)) {
ecopts.flags &= ~ECONV_UNIVERSAL_NEWLINE_DECODER;
}
if ((ecopts.flags & (ECONV_UNIVERSAL_NEWLINE_DECODER|
ECONV_CRLF_NEWLINE_ENCODER|
ECONV_CR_NEWLINE_ENCODER)) == 0) {
if (from_enc && from_enc == to_enc) { if (from_enc && from_enc == to_enc) {
return -1; return -1;
} }
@ -1895,8 +1912,6 @@ str_transcode0(int argc, VALUE *argv, VALUE *self, rb_econv_option_t *ecopts)
} }
else { else {
if (encoding_equal(from_e, to_e)) { if (encoding_equal(from_e, to_e)) {
/* newline conversion only.
* xxx: this assumes ascii compatible encoding. */
from_e = ""; from_e = "";
to_e = ""; to_e = "";
} }
@ -1908,7 +1923,7 @@ str_transcode0(int argc, VALUE *argv, VALUE *self, rb_econv_option_t *ecopts)
dest = rb_str_tmp_new(blen); dest = rb_str_tmp_new(blen);
bp = (unsigned char *)RSTRING_PTR(dest); bp = (unsigned char *)RSTRING_PTR(dest);
transcode_loop(&fromp, &bp, (sp+slen), (bp+blen), dest, str_transcoding_resize, from_e, to_e, ecopts); transcode_loop(&fromp, &bp, (sp+slen), (bp+blen), dest, str_transcoding_resize, from_e, to_e, &ecopts);
if (fromp != sp+slen) { if (fromp != sp+slen) {
rb_raise(rb_eArgError, "not fully converted, %"PRIdPTRDIFF" bytes left", sp+slen-fromp); rb_raise(rb_eArgError, "not fully converted, %"PRIdPTRDIFF" bytes left", sp+slen-fromp);
} }