GNU bug report logs - #47850
28.0.50; find-file-noselect non-nil nowarn argument effect in read-only buffers

Previous Next

Package: emacs;

Reported by: dalanicolai <at> gmail.com

Date: Sat, 17 Apr 2021 21:01:03 UTC

Severity: normal

Tags: fixed

Found in version 28.0.50

Fixed in version 28.1

Done: Lars Ingebrigtsen <larsi <at> gnus.org>

Bug is archived. No further changes may be made.

Full log


Message #8 received at 47850 <at> debbugs.gnu.org (full text, mbox):

From: Lars Ingebrigtsen <larsi <at> gnus.org>
To: dalanicolai <at> gmail.com
Cc: 47850 <at> debbugs.gnu.org
Subject: Re: bug#47850: 28.0.50; find-file-noselect non-nil nowarn argument
 effect in read-only buffers
Date: Tue, 04 May 2021 11:47:04 +0200
dalanicolai <at> gmail.com writes:

> Find-file-noselect buffer in non-existing directory with non-nil nowarn
> argument e.g:
>
> (switch-to-buffer (find-file-noselect "non-existing-dir/test.el" t))
>
> the buffer will open in read-only mode.
>
> Note that with a nowarn is nil argument, the buffer opens in an
> editable mode as expected.
>
> If it is not considered a technical bug then I would consider it a
> documentation bug.

I think it's a bug.  It stems from this thing:

(defun after-find-file (&optional error warn noauto
				  _after-find-file-from-revert-buffer
				  nomodes)
[...]
	    (cond
	     ((not warn) nil)
	     ((and error (file-exists-p buffer-file-name))
	      (setq buffer-read-only t)
	      "File exists, but cannot be read")
	     ((and error (file-symlink-p buffer-file-name))
	      "Symbolic link that points to nonexistent file")
	     ((not buffer-read-only)
	      (if (and warn

Already here the code is pretty dubious, because we've already skipped
all this if warn is non-nil...  Anyway, here's the problem:

[...]

	     (t
	      (setq buffer-read-only nil)
	      (unless (file-directory-p default-directory)
		"Use M-x make-directory RET RET to create the directory and its parents")))))

We should do this action even if warn is non-nil, I think?  Or more
generally, we should do all the actions in that cond, but not actually
issue the warning.

I've now done this on the trunk...  this means that those file-exists-p
things are run in the nil WARN case, so the behaviour is slightly
different than before, and it's possible that this may lead to
regressions, I think, but it seems unlikely to me.

-- 
(domestic pets only, the antidote for overdose, milk.)
   bloggy blog: http://lars.ingebrigtsen.no




This bug report was last modified 4 years and 19 days ago.

Previous Next


GNU bug tracking system
Copyright (C) 1999 Darren O. Benham, 1997,2003 nCipher Corporation Ltd, 1994-97 Ian Jackson.