Unescaping character display.

Unescaping character display.

[rvclayton]

emacs 24.5.1 is displaying some characters as numeric escapes.  For example,
the love-minor-mode description in the list-packages command output is "Minor
mode for working on L\303\226VE projects" (minus the double quotes).  My init
file contains (from dotemacs.de/dotfiles/BenoitGoudreault-Emond.emacs.html)

  (set-language-environment "Latin-1")
  (set-terminal-coding-system 'iso-latin-1)
  (setq unibyte-display-via-language-environment t)

What can I do to get the such characters to display in their unescaped form?

Re: Unescaping character display.

[John Ankarström]

rvclayton@acm.org writes:

> For example, the love-minor-mode description in the
> list-packages command output is "Minor mode for working on
> L\303\226VE projects" (minus the double quotes).

If it's any help, I can tell you that that letter in the middle
of L and VE is Ö - the same as in my name! :-)

On my installation of Emacs, (describe-package 'love-minor-mode)
displays the following summary:

> Minor mode for working on LÖVE projects

I think it's supposed to be jokingly pronounced with an
"ea"-sound as in "early" :-)

- John

Re: Unescaping character display.

[Eli Zaretskii]

> From: rvclayton@acm.org
> Date: Sun, 28 May 2017 12:06:24 -0400
> 
> emacs 24.5.1 is displaying some characters as numeric escapes.  For example,
> the love-minor-mode description in the list-packages command output is "Minor
> mode for working on L\303\226VE projects" (minus the double quotes).  My init
> file contains (from dotemacs.de/dotfiles/BenoitGoudreault-Emond.emacs.html)
> 
>   (set-language-environment "Latin-1")
>   (set-terminal-coding-system 'iso-latin-1)
>   (setq unibyte-display-via-language-environment t)
> 
> What can I do to get the such characters to display in their unescaped form?

Delete the above 3 lines, restart Emacs, and see if you get a more
reasonable display.

(Those 3 lines are remnants of issues long gone, especially the 3rd
one.  You shouldn't need them.)

Re: Unescaping character display.

[R. Clayton]

Thanks for your reply to my message.

  If it's any help, I can tell you that that letter in the middle
  of L and VE is Ö - the same as in my name! :-)

Yes, I know: the (two different typefaces of) umlaut-o shows up in your
message, which I'm reading in rmail in emacs.  But running list-packages here
still gives me \303\226.  I was thinking it might be a font problem, but I
guess it's more of an encoding problem.

Re: Unescaping character display.

[rvclayton]

Thanks for replying to my message.

  Delete the above 3 lines, restart Emacs, and see if you get a more
  reasonable display.

They're gone, but the problem remains.

Re: Unescaping character display.

[Eli Zaretskii]

> From: rvclayton@acm.org
> Date: Mon, 29 May 2017 10:39:49 -0400
> 
> Thanks for replying to my message.
> 
>   Delete the above 3 lines, restart Emacs, and see if you get a more
>   reasonable display.
> 
> They're gone, but the problem remains.

What happens in "emacs -Q"?

Re: Unescaping character display.

[R. Clayton]

  What happens in "emacs -Q"?

The same problem after futzing with package-archives.  \303\226 is the utf-16
octal escape sequence for a capital O with an umlaut (or diaeresis if you want
to be that way).  I think I'll have to figure out where to drop a few utf-16
indicators into my init files.

Re: Unescaping character display.

[Eli Zaretskii]

> From: rvclayton@acm.org (R. Clayton)
> Date: Mon, 29 May 2017 15:34:11 -0400
> 
>   What happens in "emacs -Q"?
> 
> The same problem after futzing with package-archives.  \303\226 is the utf-16
> octal escape sequence for a capital O with an umlaut (or diaeresis if you want
> to be that way).

You mean UTF-8, not UTF-16.

In any case, at this point you should file a bug report with all the
details, because "emacs -Q" shouldn't fail to decode UTF-8 text.

> I think I'll have to figure out where to drop a few utf-16
> indicators into my init files.

??? Can't say I understand what that is about.  "utf-16 indicators"?
what's that?

Re: Unescaping character display.

[R. Clayton]

  You mean UTF-8, not UTF-16.

Yeah, probably.

  In any case, at this point you should file a bug report with all the
  details, because "emacs -Q" shouldn't fail to decode UTF-8 text.

That's what I was thinking (or "thinking").  I'd seen o-with-umlauts in rmail,
so I figured utf-8 was covered correctly, leaving utf-16 (or utf-32, but that's
ridiculous).

  Can't say I understand what that is about.  "utf-16 indicators"?  what's
  that?

A symbol or string or keyword indicating utf-16 is relevant in some way, like
is used in gnus-group-name-charset-group-alist.  I imagined such a mechanism
exists to solve the problem, but a quick look around wasn't too promising.

I'll dink around with this a while longer.  If I don't get anywhere, I'll file
a bug report.  Thanks for your help.