Re: cp1251 coding system and files

[Kenichi Handa]

In article <address@hidden>, Ognyan Kulev <address@hidden> writes:

> Kenichi Handa wrote:
>>  In short, Emacs 22 provides only bulgarian-bdf and
>>  bulgarian-phonetic input methods for Bulgarian.  They are
>>  listed when you type C-h L RET
>>  (describe-language-environment).  They produce characters of
>>  charset mule-unicode-0100-24ff, and thus can be saved safely
>>  by cp1251.
>>  
>>  By the way, Emacs doesn't have input methods
>>  cyrillic-bulgarian-*.

> It turns out that I was using my Bulgarian input methods for Emacs 21.
> I apologize for bothering -- it looks like I was too excited when I
> found this "bug".

No need of apologizing.  By your report, I've just realized
this:

Coding system cp1251 of Emacs 21 handles characters of
Emacs' internal charset cyrillic-iso8859-5 (thus it can
read/write only characters belonging to iso8859-5), but that
of Emacs 22 handles characters of Emacs' internal charset
mule-unicode-0100-24ff (thus it can read/write all cp1251
characters).

The reason why you can use those input methods in Emacs 21
is that they produce cyrillic-iso8859-5 characters.  Right?

Hmmm, it seems that this is a regression of Emacs 22.  I'll
find a way to solve that.  Please wait for a while.

---
Kenichi Handa
address@hidden