GNU bug report logs - #74511
31.0.50; `dictionary-search-interface' overrides the user's customization of `dictionary-read-word-function'

Previous Next

Package: emacs;

Reported by: Steven Allen <steven <at> stebalien.com>

Date: Sun, 24 Nov 2024 18:57:01 UTC

Severity: normal

Found in version 31.0.50

Fixed in version 30.1

Done: Eshel Yaron <me <at> eshelyaron.com>

Bug is archived. No further changes may be made.

Full log


View this message in rfc822 format

From: Eshel Yaron <me <at> eshelyaron.com>
To: Eli Zaretskii <eliz <at> gnu.org>
Cc: steven <at> stebalien.com, 74511 <at> debbugs.gnu.org
Subject: bug#74511: 31.0.50; `dictionary-search-interface' overrides the user's customization of `dictionary-read-word-function'
Date: Fri, 29 Nov 2024 07:30:35 +0100
Hi,

Eli Zaretskii <eliz <at> gnu.org> writes:

>> From: Eshel Yaron <me <at> eshelyaron.com>
>> Cc: Eli Zaretskii <eliz <at> gnu.org>,  74511 <at> debbugs.gnu.org
>> Date: Mon, 25 Nov 2024 22:27:57 +0100
>> 
>> Steven Allen <steven <at> stebalien.com> writes:
>> 
>> > Is adding the following to `dictionary-search-interface' not sufficient?
>> >
>> >     :initialize 'custom-initialize-changed
>> >
>> > That would apply the `dictionary-search-interface' setting if and only
>> > if specified by the user while still alowing the user to, e.g., use the
>> > customize interface.
>> 
>> Ah yes, seems even better than my patch since it also makes it possible
>> to set dictionary-search-interface before loading dictionary.el, thanks!
>> 
>> Eli, does this seem OK for the release branch?
>> 
>> diff --git a/lisp/net/dictionary.el b/lisp/net/dictionary.el
>> index 58c2e9771ba..42fb8c57b40 100644
>> --- a/lisp/net/dictionary.el
>> +++ b/lisp/net/dictionary.el
>> @@ -317,6 +317,7 @@ dictionary-search-interface
>>                        dictionary-read-dictionary-function)
>>                       vals))
>>           (set-default-toplevel-value symbol value))
>> +  :initialize #'custom-initialize-changed
>>    :version "30.1")
>
> Don't you mean custom-initialize-set?

No, I don't: custom-initialize-set calls the :set function if the symbol
does not have a default-toplevel-value, which is when the option is
initialized while loading dictionary.el.  So using this :initialize
function does not fix the recipe Steve provided in the original post.

> If I'm missing something, please describe what will happen in the
> various scenarios when dictionary-search-interface and the other
> related options are customized, whether by setq or via Custom, and
> when they are not customized the first time dictionary.el is loaded.
> When we understand all the effects of this, we can reason whether the
> change is safe for the emacs-30 branch (which it better be, since
> these options were introduced in Emacs 30, AFAIR).

With custom-initialize-changed we get the following:

- If you customize the other options but not dictionary-search-interface
  before loading dictionary.el, then your settings persist (they are not
  overridden) after loading dictionary.el.  If you later customize
  dictionary-search-interface, then it affects does the other options,
  as expected.
- If you do customize dictionary-search-interface before loading
  dictionary.el, then it does set the other options when you load
  dictionary.el, whether or not you've also set the other options (so it
  doesn't make sense to customize both dictionary-search-interface and
  the other options, you either customize dictionary-search-interface or
  individual options).

So AFAICT it's safe in the sense that it facilitates use cases that
currently do not work (customizing the other options before loading
dictionary.el) without prohibiting any existing use case.


Best,

Eshel




This bug report was last modified 211 days ago.

Previous Next


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