On Sun, Oct 13, 2002 at 01:12:53AM +0200, Miloslav Trmac wrote:
> Hello,
> the linux kernel keyboard handling code currently has compose table
> with result 'unsigned char', wich means that compose/dead keys can't
> produce UTF-8 characters. I'm willing to make a patch both to
> kernel and kbd. Since this means adding a new ioctl, I'd like to ask
> whether you want any other interface changes in this KD[GS]KDIACR2
> other than just widening the 'result' in struct kbdiacr.
Sounds reasonable to me.
In the kbd TODO file:
- selection can now cut & paste 8-bit chars, but not yet Unicode
- compose does not work together with Unicode
(struct kbdiacr has three chars: base + diacr -> result,
but the result should be a string)
- strings produced by a function key cannot contain \000;
many other places this char is excluded
- current status (translation table, Unicode output) can be set,
but not read out from the kernel; thus, showconsolefont changes the
status, but cannot restore it
So, this new KD[GS]KDIACR2, or let us say KD[GS]KCOMPOSE,
should return a string, not a single Unicode character.
(Think of cases where you want an accented character
but the precomposed version does not exist in Unicode.
Unicode avoids precomposed versions, and has them only
when some existing character set already had them.)
> And BTW, can this get in the "stable" 2.4?
It should be possible given there is a serious need for this.
Do 2.5 first.
Andries
Let me cc linux-utf8, there live a lot of people
that may have comments.
--
Linux-UTF8: i18n of Linux on all levels
Archive: http://mail.nl.linux.org/linux-utf8/
