bug#81606: Mojibake with UTF-8 file names

Eli Zaretskii <[email protected]>
Newsgroups gmane.emacs.bugs
Message-ID <[email protected]>
> Date: Sun, 16 Aug 2026 14:29:54 +0000
> Cc: [email protected], [email protected]
> From: Werner LEMBERG <[email protected]>
> 
> >> If a master file `foo` is specified in the local variables block,
> >> AUCTeX checks whether it is exists, otherwise it gets created.
> >> This string `foo` must be given in the encoding specified by the
> >> `coding` cookie.
> > 
> > But this is trivially correct: any non-ASCII text in a file which has
> > a 'coding:' cookie _must_ be encoded in the encoding specified by
> > 'coding:".  Right?
> 
> There might be situations where the file encoding and the file name
> encoding differ, for example on Windows.

No, if the file name is specified by the same file that has the
'coding:' cookie, that cannot happen.

> I don't know enough about the encoding handling in Emacs whether it
> automatically maps UTF-8 encoded file name strings (as implied by
> the `coding` cookie) to UTF-16...

Of course it does.  How else could Emacs support arbitrary non-ASCII
characters in file names on Windows?

> > If you want to require the 'coding:' cookie in all cases, that's
> > fine, but then latexenc.el is not needed at all, see above.
> 
> Well, the AUCTeX documentation doesn't require it yet — otherwise I
> wouldn't have noticed the problem.  I think it's a good thing if Emacs
> has a heuristic fallback mechanism to handle the case without a
> `coding` cookie.

And for that case I think the recommendation should be to encode the
file name in UTF-8.
lmpx.com only provides a reader for public news (NNTP) servers. It is not affiliated with the servers or forums shown here and is not responsible for the content of articles, which is written by their respective authors.