prefer-coding-system, utf-8, iso-8859-2

prefer-coding-system, utf-8, iso-8859-2

[Emil Perhinschi]

My problem:

I have received a file coded in utf-8. I have to translate parts
of it into Romanian

prefer-coding-system utf-8

C-x C-f filename

echo $LANG=ro_RO

set-input-method latin-2-postfix

Then I write my part.


Trying to save:

C-x C-s

the result is :

These default coding systems were tried:
  mule-utf-8-unix
However, none of them safely encodes the target text.

Select one of the following safe coding systems:
  x-ctext iso-2022-7bit raw-text emacs-mule no-conversion
  ctext-no-compositions iso-2022-8bit 

etc.



Yes, I know there was a thread in May covering exactly this
topic, but the suggested solution, prefer-coding-system, does not
work for me.

Emacs version: 

GNU Emacs 21.2.1 (i386-debian-linux-gnu, X toolkit, Xaw3d scroll
bars) of 2002-03-22 on raven, modified by Debian


Is there any solution to this? Am I doing something wrong?

thank you,

Emil Perhinschi

Re: prefer-coding-system, utf-8, iso-8859-2

[Eli Zaretskii]

> Date: Wed, 17 Sep 2003 12:51:41 +0300
> From: Emil Perhinschi <emilper@home.ro>
> 
> prefer-coding-system utf-8
> 
> C-x C-f filename

After these two lines, what does Emacs print if you say

  M-: buffer-file-coding-system RET

?

> set-input-method latin-2-postfix

I think this is your problem: this input method produces iso-8859-2
characters, not Unicode characters (Emacs 21.x doesn't know that they
are the same).  You need to use a Unicode input method, like TeX.  Or
install some add-on package such as ucs-tables, and turn on
unification on encoding.