On Monday 25 April 2005 01:41 pm, Eugeniy Meshcheryakov wrote:
> 25 квітня 2005 о 13:08 -0400 Daniel Burrows написав(-ла):
> > I suppose that the gettext-friendly solution is something like
> >
> > char ch=toupper(response[0]);
> >
> > if(ch == _("Yes")[0])
> > ...
>
> Bad idea. When using UTF-8 locale, and if aptitude will be translated to
> Ukrainian (for example), we will have:
> if(ch == "Так"[0])
> This will be true for all (or most of) cyrillic letters. Т = "\x0d\xb0"
> and all cyrillic letters have first byte \x0d.
Right now aptitude is utterly broken in UTF-8 anyway. When the UTF-8
conversion is done, it should handle input in UTF-8 properly as well (or
everything will break, as you pointed out).
Daniel
--
/----------------- Daniel Burrows <d.burrows4@verizon.net> -----------------\
| Somewhere, just out of sight, the unicorns are gathering. |
\--------------------- A duck! -- http://www.python.org --------------------/
Attachment:
pgpi7Qunalot3.pgp
Description: PGP signature