[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: debconf and locale specific characters


> > It is save to assume that the data from debconf is in the charset of the
> > current locale? Then foo=`echo $RET|iconv -t UTF8` after the db_get
> > would suffice. Of course the locale could change between config and
> > postinst so maybe I should convert to UTF-8 and store the result in the
> > debconf db in the config script?
> The one thing I would worry about is checking the encoding of data entered
> using the GNOME frontend, which may have different behavior than the text
> frontends.  For the text frontends, I certainly think it's safe to assume
> that the input data arrives in the locale charset.
> I agree there's a (slim) chance of the charset changing between config and
> postinst, so recoding in the config script itself is best.

I've got a gut feeling that debconf should be the one who
handles the encoding, because packages won't be able to know the 
full context.

Would it be better to force all debconf frontends to have 
sanitized UTF-8 data?

This will probably break quite a few things due to the fact that we've 
never done this before,..


Junichi Uekawa, Debian Developer
17D6 120E 4455 1832 9423  7447 3059 BF92 CD37 56F4

Reply to: