Re: IPA characters from subprocesses

[Kai Großjohann]

dk2993@accela.net writes:

>   Here it is for the missing-char box:
>
>   character: ɛ (01210733, 332251, 0x511db)
>     charset: mule-unicode-0100-24ff

Ah.  Well, Emacs often considers like-looking characters to be
different, if different encodings are used.  So the character you
type is in one encoding and the character read from the file is in
another encoding.

There are facilities for unifying some parts of Unicode with the
corresponding non-Unicode encodings, see unify-8859-on-encoding-mode
(on by default in Emacs 21.3), unify-8859-on-decoding-mode, and also
utf-translate-cjk.

I'm afraid that the characters you're talking about might not be
mentioned in any of those places, though.  Hm.

But what is really surprising is that Emacs can obviously convert the
character between the encodings, as it works to send the IPA
character to the file using utf-8 encoding.  Hm.  Only the other way
round it does not seem to work.  Hm.

Maybe it is easiest to just install a comprehensive Unicode font so
that you can see the characters.  However, this will still suffer
from the problem that you will have different characters in the
buffer that look alike, so C-s will produce surprising results,
missing some of the characters.

You could look at ucs-tables.el and augment the tables found there by
the IPA characters, or add a new table for the IPA characters.  WDYT?
-- 
~/.signature