You are viewing a plain text version of this content. The canonical link for it is here.
Posted to users@subversion.apache.org by Eric Lemes <er...@gmail.com> on 2006/06/08 21:41:04 UTC
Problems Encoding
Hello there,
I'm trying to get more information about how svn Windows command line tools
works with encoding.
When I do a
svn log http://myrepo -r x
I got allways the standard output in the machine's codepage?
My problem is... I'm trapping the stdout of svn tools in a .Net application.
When I do a --xml in svn command, I receive it in UTF8. When I do svn or
svnlook without --xml option, I got wrong chars trapping the output in UTF8
or ANSI.
Is this a problem in brazilian localization? Anyone knows how to tell svn
tools to everytime sends his output in utf-8 or to force svn to US locale?
[]'s
Eric Lemes
Re: Problems Encoding
Posted by Mathias Weinert <Ma...@hyposystems.de>.
Eric Lemes wrote:
> On 6/9/06, Samuel Langlois <sl...@ilog.fr> wrote:
> Hello,
>
> I have the same thing in French localization:
> C:\>svn --version
> svn, version 1.3.2 (r19776)
> compil?\195?\169 May 26 2006, 13:10:00
>
> It seems svn spits UTF-8, which the basic Windows Command Prompt cannot
handle.
> Setting the LANG environment variable to en_US switches to English
messages, which is an acceptable workaround for me.
>
> Well.. I think he doesn't spit UTF-8. I'm trapping the stdout with a C#
app, and reading the input with utf-8 encoding for svn --xml, I got all
chars okay. Trapping svnlook stdout I got scrambled chars.
>
> I think svn command line tools uses Windows Regional settings for his
localization and he spits chars in the "codepage" configured in Regional
Settings "Advanced" Tab (in the "select a language for non-unicode
programs). My problem is that when I set this to "Brazilian Portuguese",
svn tools spit good chars in windows console, but I can't trap the stdout
from my C# app in UTF8 or Default (ANSI Encoding, iso-8859-1 I think).
>
> The LANG=en_US didn't affected anything.
>
> I did a test with a common text-file, saved as ANSI with weird chars:
"çã". In a HEX editor, these to chars goes as E7 and E3.
>
> With a svnlook > textfile.txt (as ANSI too), I got 87 and C6 for the
same chars.
>
> Seeing the output of the UTF-8 file (parsed from svn --xml), I got two
bytes for every weird char: "ç" = C3 A7, "ã" = C3 A3.
>
>
> Thanks anyway,
Just set the APR_ICON_PATH environment variable to the iconv
path of your subversion installation and it should work.
Mathias
---------------------------------------------------------------------
To unsubscribe, e-mail: users-unsubscribe@subversion.tigris.org
For additional commands, e-mail: users-help@subversion.tigris.org
Re: Problems Encoding
Posted by Eric Lemes <er...@gmail.com>.
On 6/9/06, Samuel Langlois <sl...@ilog.fr> wrote:
>
> Hello,
>
> I have the same thing in French localization:
> C:\>svn --version
> svn, version 1.3.2 (r19776)
> compil?\195?\169 May 26 2006, 13:10:00
>
> It seems svn spits UTF-8, which the basic Windows Command Prompt cannot
> handle.
> Setting the LANG environment variable to en_US switches to English
> messages, which is an acceptable workaround for me.
>
Well.. I think he doesn't spit UTF-8. I'm trapping the stdout with a C# app,
and reading the input with utf-8 encoding for svn --xml, I got all chars
okay. Trapping svnlook stdout I got scrambled chars.
I think svn command line tools uses Windows Regional settings for his
localization and he spits chars in the "codepage" configured in Regional
Settings "Advanced" Tab (in the "select a language for non-unicode
programs). My problem is that when I set this to "Brazilian Portuguese", svn
tools spit good chars in windows console, but I can't trap the stdout from
my C# app in UTF8 or Default (ANSI Encoding, iso-8859-1 I think).
The LANG=en_US didn't affected anything.
I did a test with a common text-file, saved as ANSI with weird chars: "çã".
In a HEX editor, these to chars goes as E7 and E3.
With a svnlook > textfile.txt (as ANSI too), I got 87 and C6 for the same
chars.
Seeing the output of the UTF-8 file (parsed from svn --xml), I got two bytes
for every weird char: "ç" = C3 A7, "ã" = C3 A3.
Thanks anyway,
[]'s
Eric Lemes
RE: Problems Encoding
Posted by Samuel Langlois <sl...@ilog.fr>.
Hello,
I have the same thing in French localization:
C:\>svn --version
svn, version 1.3.2 (r19776)
compil?\195?\169 May 26 2006, 13:10:00
It seems svn spits UTF-8, which the basic Windows Command Prompt cannot handle.
Setting the LANG environment variable to en_US switches to English messages, which is an acceptable workaround for me.
Hope this helps,
--
Samuel Langlois
________________________________________
From: Eric Lemes [mailto:ericlemes@gmail.com]
Sent: jeudi 8 juin 2006 23:41
To: users@subversion.tigris.org
Subject: Problems Encoding
Hello there,
I'm trying to get more information about how svn Windows command line tools works with encoding.
When I do a
svn log http://myrepo -r x
I got allways the standard output in the machine's codepage?
My problem is... I'm trapping the stdout of svn tools in a .Net application. When I do a --xml in svn command, I receive it in UTF8.
When I do svn or svnlook without --xml option, I got wrong chars trapping the output in UTF8 or ANSI.
Is this a problem in brazilian localization? Anyone knows how to tell svn tools to everytime sends his output in utf-8 or to force
svn to US locale?
[]'s
Eric Lemes
---------------------------------------------------------------------
To unsubscribe, e-mail: users-unsubscribe@subversion.tigris.org
For additional commands, e-mail: users-help@subversion.tigris.org