auctex-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Slash at the end of directory (was Re: oddity with TeX-style-private ?)


From: Ikumi Keita
Subject: Slash at the end of directory (was Re: oddity with TeX-style-private ?)
Date: Sat, 24 Oct 2020 22:18:17 +0900

>>>>> Tassilo Horn <tsdh@gnu.org> writes:
> Uwe Brauer <oub@mat.ucm.es> writes:
>>>>> "CB" == Colin Baxter <m43cap@yandex.com> writes:
CB> However, the default setting for this variable, given in
CB> 'tex-site.el' is "/usr/local/var/auctex", which does not end in '/'.
 
Uwe> Right, this also crossed my mind and I tried it out.
Uwe> The '/' did not make any difference. I was also wondering.

Tassilo> Probably a relict of ancient times where paths were build using
Tassilo> `concat' rather than `expand-file-name' and friends, I don't
Tassilo> know...

The commit [1] removed most of the occurences of "Must end with a
directory separator." The one discussed here must be just a remnant of
them. I think that statement can be deleted safely.

According to the words from David in [2], it seems that the removal was
done for the sake of xemacs.

By the way, I think that file-name-as-directory should be re-enabled for
TeX-macro-{global,private}, considering that we no longer support
xemacs, because it seems that TeX-strip-extension and TeX-parse-path
assume that those variables have list of strings which end with trailing
slashes.
TeX-strip-extension:
  (let* (...
         (dir (expand-file-name (or (file-name-directory strip) "./"))))
         ...
            (member dir TeX-macro-global)
            (member dir TeX-macro-private))

TeX-parse-path:
      (setq entry (file-name-as-directory
                   (if (string-match "/?/?\\'" entry)
                       (substring entry 0 (match-beginning 0))
                     entry)))
      ...
          (member entry (append '("/" "\\") TeX-macro-global))

Regards,
Ikumi Keita

[1] 
https://git.savannah.gnu.org/cgit/auctex.git/commit/?id=4ac002797d84dcb57ee4ec10c7e120ad71bf0c69
[2] https://debbugs.gnu.org/cgi/bugreport.cgi?bug=31150
"XEmacs used to do things differently (and wasn't tolerant about it as
opposed to Emacs if I remember correctly) and thus many packages used
that behavior."



reply via email to

[Prev in Thread] Current Thread [Next in Thread]