Re: Problems with Cyrillic encoding

[Kenichi Handa <handa@m17n.org>]

In article <20030217125917.1F5337C018@mail.filanet.dk>, Claus Tondering <ct@filanet.dk> writes:
> In GNU Emacs 21.2.1 (i686-pc-linux-gnu, X toolkit, Xaw3d scroll bars)
[...]
> I can type Russian (Cyrillic) letters in two ways:

> * If I select a Russian keyboard and type Russian letters, they get
>   into the Emacs buffer with character values in the range
>   0x51450-0x5146f (for the lower case letters).

>   Calling the the built-in Lisp function
>   find-coding-systems-region-internal on these characters returns
>   mule-utf-8 as a possible encoding.

> * If I use a Latin keyboard and set the input method to
>   cyrillic-jcuken, the Russian letters are inserted into the Emacs
>   buffer as characts with values in the range 0xe50-0xe6f (for the
>   lower case letters).

>   Calling find-coding-systems-region-internal on these characters does
>   *not* return mule-utf-8 as a possible encoding.

> This has bad consequences when I try to use Russian characters in
> email and Gnus, because UTF-8 is not chosen as a valid encoding in the
> last case.

Please try the latest pretest version 21.2.95 available at
<ftp://alpha.gnu.org/gnu/emacs/pretest/>.

With this version, both kind of Russian characters
(0x51450-0x5146f and 0xe50-0xe6f) can be encoded by both
utf-8 and iso-8859-5.

But, I don't know why characters of the range
0x51450-0x5146f is input by a Russian keyboard.

As something about X's keysym handling is changed, the
behaviour of the pretest version may be different from 21.2.

---
Ken'ichi HANDA
handa@m17n.org