GNU bug report logs - #59486
completion-auto-wrap disobeyed by vertical navigation

Previous Next

Package: emacs;

Reported by: Juri Linkov <juri <at> linkov.net>

Date: Tue, 22 Nov 2022 17:46:01 UTC

Severity: normal

Fixed in version 30.0.50

Done: Juri Linkov <juri <at> linkov.net>

Bug is archived. No further changes may be made.

Full log


Message #44 received at 59486 <at> debbugs.gnu.org (full text, mbox):

From: Juri Linkov <juri <at> linkov.net>
To: Eli Zaretskii <eliz <at> gnu.org>
Cc: 59486 <at> debbugs.gnu.org
Subject: Re: bug#59486: completion-auto-wrap disobeyed by vertical navigation
Date: Sun, 05 Nov 2023 19:53:23 +0200
>> The argument zero to get-buffer-window AFAIU means that it will return
>> non-nil when the buffer is shown in some window on an iconified frame,
>> and I wonder why we would consider such a buffer "visible".
>
> (get-buffer-window "*Completions*" 0) is used everywhere in minibuffer.el
> and simple.el, so the argument zero is for compatibility with other
> minibuffer completion commands.

During testing I discovered that the condition should be more complex.
The problem is that the logic should be bound only to the minibuffer
that showed the completions, not in other minibuffers within a set of
recursive minibuffers.  An example: `M-x TAB C-h v down RET'
raised an error "Minibuffer is not active for completion".
This error comes from `choose-completion-string',
so I copied the same logic from `choose-completion-string':

  (when-let ((window (get-buffer-window "*Completions*" 0)))
    (when (eq (buffer-local-value 'completion-reference-buffer
                                  (window-buffer window))
              (window-buffer (active-minibuffer-window)))
      cmd))

>> Same here.
>
> Here is a new patch:

Now pushed the fixed patch.




This bug report was last modified 1 year and 192 days ago.

Previous Next


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