Re: term, utf-8 and cooked mode, combining characters

[Marcus Brinkmann]

On Mon, Sep 16, 2002 at 12:31:44PM -0700, Thomas Bushnell, BSG wrote:
> Marcus Brinkmann <Marcus.Brinkmann@ruhr-uni-bochum.de> writes:
> 
> > The other related question is about combining characters.  Should backspace
> > delete the base character plus all combining characters, or only the last
> > combining character?  This affects term and the console server.
> 
> If you are doing real Unicode, they suggest that it should delete the
> last combining character.  But this is of course different than what
> you get with a simple implementation of Latin-1.

I am a bit at a loss here.  I am not doing "real" Unicode, as for example I
didn't implement surrogates, or right to left writing.  But I am not doing
Latin-1 only either.

I am using the UTF-8 locale with a lot of applications right now very
successfully (hurra, I can read all my spam, even korean).  I can input
unicode characters.  OTOH, with just a simple configuration change (console
server's --encoding option), I can get a console in any encoding supported
by iconv (like, ISO8859-1).

Now, so far we don't support combining characters, but as I previously said
I think I can add that without too much hassles.

Anyway, the problem is that term is not multi-byte aware.  OTOH, I was not
really using term, but libreadline in bash, but both have the same problems. 
A newer libreadline (4.3) is multi-byte aware, and I will try out how it
behaves.  I think we should do the same in term.  In fact, I remember that
integrating libreadline in term is a TODO list item, and it would make a lot
of sense to just punt this to libreadline.  Do you think using libreadline
in term's cooked mode is difficult?

Thanks,
Marcus

-- 
`Rhubarb is no Egyptian god.' GNU      http://www.gnu.org    marcus@gnu.org
Marcus Brinkmann              The Hurd http://www.gnu.org/software/hurd/
Marcus.Brinkmann@ruhr-uni-bochum.de
http://www.marcus-brinkmann.de/