moparisthebest/wget
1
0
Fork 0
mirror of https://github.com/moparisthebest/wget synced 2024-07-03 16:38:41 -04:00
Code Issues Releases Wiki Activity

Fix problem with IDN and UTF-8 encoding.

Browse Source
This commit is contained in:
Merinov Nikolay 2011-07-26 09:27:08 +02:00 committed by Giuseppe Scrivano
parent 8c69d9f4a9
commit 37c85ecba3
6 changed files with 32 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

6
src/ChangeLog
View File

@ -1,3 +1,9 @@
2011-07-20 Merinov Nikolay <kim.roader@gmail.com>
* iri.c (remote_to_utf8): Add test for non-ASCII symbols with
UTF-8 URI encoding.
* res.c (res_retrieve_file): Fix url_parse call.
2011-06-08 Giuseppe Scrivano <gscrivano@southpole.se> 2011-06-08 Giuseppe Scrivano <gscrivano@southpole.se>
* retr.c (retrieve_from_file): Parse the url careless if IRI is enabled. * retr.c (retrieve_from_file): Parse the url careless if IRI is enabled.

15
src/iri.c
View File

@ -264,6 +264,21 @@ remote_to_utf8 (struct iri *i, const char *str, const char **new)
if (!i->uri_encoding) if (!i->uri_encoding)
return false; return false;
/* When `i->uri_encoding' == "UTF-8" there is nothing to convert. But we must
test for non-ASCII symbols for correct hostname processing in `idn_encode'
function. */
if (!strcmp (i->uri_encoding, "UTF-8"))
{
int i, len = strlen (str);
for (i = 0; i < len; i++)
if ((unsigned char) str[i] >= (unsigned char) '\200')
{
*new = strdup (str);
return true;
}
return false;
}
cd = iconv_open ("UTF-8", i->uri_encoding); cd = iconv_open ("UTF-8", i->uri_encoding);
if (cd == (iconv_t)(-1)) if (cd == (iconv_t)(-1))
return false; return false;

2
src/res.c
View File

@ -552,7 +552,7 @@ res_retrieve_file (const char *url, char **file, struct iri *iri)
opt.timestamping = false; opt.timestamping = false;
opt.spider = false; opt.spider = false;
url_parsed = url_parse (robots_url, &url_err, iri, true); url_parsed = url_parse (robots_url, &url_err, i, true);
if (!url_parsed) if (!url_parsed)
{ {
char *error = url_error (robots_url, url_err); char *error = url_error (robots_url, url_err);

6
tests/ChangeLog
View File

@ -1,3 +1,9 @@
2011-06-03 Merinov Nikolay <kim.roader@gmail.com>
* Test-idn-cmd-utf8.px: Added test for idn with utf-8 local encoding.
* Test-idn-robots-utf8.px: Added test for idn with utf-8 local encoding
and robots.txt file.
* Makefile.am, run-px: Add new tests.
2011-04-19 Giuseppe Scrivano <gscrivano@gnu.org> 2011-04-19 Giuseppe Scrivano <gscrivano@gnu.org>
* Makefile.am (LIBS): Add $(LIB_CLOCK_GETTIME). * Makefile.am (LIBS): Add $(LIB_CLOCK_GETTIME).

2
tests/Makefile.am
View File

@ -90,7 +90,9 @@ EXTRA_DIST = FTPServer.pm FTPTest.pm HTTPServer.pm HTTPTest.pm \
Test-idn-headers.px \ Test-idn-headers.px \
Test-idn-meta.px \ Test-idn-meta.px \
Test-idn-cmd.px \ Test-idn-cmd.px \
Test-idn-cmd-utf8.px \
Test-idn-robots.px \ Test-idn-robots.px \
Test-idn-robots-utf8.px \
Test-iri.px \ Test-iri.px \
Test-iri-percent.px \ Test-iri-percent.px \
Test-iri-disabled.px \ Test-iri-disabled.px \

2
tests/run-px
View File

@ -43,7 +43,9 @@ my @tests = (
'Test-idn-headers.px', 'Test-idn-headers.px',
'Test-idn-meta.px', 'Test-idn-meta.px',
'Test-idn-cmd.px', 'Test-idn-cmd.px',
'Test-idn-cmd-utf8.px',
'Test-idn-robots.px', 'Test-idn-robots.px',
'Test-idn-robots-utf8.px',
'Test-iri.px', 'Test-iri.px',
'Test-iri-percent.px', 'Test-iri-percent.px',
'Test-iri-disabled.px', 'Test-iri-disabled.px',