[svn.haxx.se] · SVN Dev · SVN Users · SVN Org · TSVN Dev · TSVN Users · Subclipse Dev · Subclipse Users · this month's index

Re: [PATCH] Subversion UTF-8 internals & gettext

From: Branko Čibej <brane_at_xbc.nu>
Date: 2004-04-20 09:26:57 CEST

Erik Huelsmann wrote:

>>--On Sunday, April 18, 2004 11:50 PM +0200 Erik Huelsmann
>><e.huelsmann@gmx.net> wrote:
>>
>>
>>
>>>From the following paragraph in the gettext manual I conclude that we
>>>
>>>
>>should
>>
>>
>>>be applying the patch below to make sure our internals stay in UTF-8.
>>>
>>>
>>I'm sort of -0 as this is one of the GNU-isms that isn't supported by most
>>non-GNU gettext. Is there a real need (i.e. something is broken) that
>>mandates us having this? -- justin
>>
>>
>
>Not really, except that output from _() and dgettext() is passed to
>svn_error_create(f). We would have to do one of two things to get consistent
>errors again:
>1) recode the output from (d)gettext ourselves to UTF-8
>or
>2) recode error text from the server to the local locale on reception.
>
>I think that option number (2) gets messy and it will become unpredictable
>which encoding is passsed along to which function.
>
>So I'd like to either bind the codeset to tell gettext to do the recoding,
>or to do it ourselves. The last option introduces one extra recoding step
>(<mo-input-encoding> -> local encoding -> utf8 -> local encoding) instead of
>(<mo-input-encoding> -> utf8 -> local encoding). Which could even be reduced
>one more step by generating utf8 .mo files.
>
>
Is there no way to (portably) tell gettext to leave well enough alone
and not do _any_ recoding at all?

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Tue Apr 20 09:27:01 2004

This is an archived mail posted to the Subversion Dev mailing list.

This site is subject to the Apache Privacy Policy and the Apache Public Forum Archive Policy.