bug#44016: 28.0.50; Add new "gnus-search" search interface to Gnus

[Eric Abrahamsen]

Stefan Monnier <monnier@iro.umontreal.ca> writes:

>> (the "(consp (cdr (completion-all-completions" bit above definitely
>> felt like I was holding the tool upside down),
>
> It's just another way (more efficient) way to write
>
>    (< 1 (length (completion-all-completions ...)))
>
> so it looks fine from where I stand.

The return value for multiple matches is not a proper list (looks like
("subject" "since" . 0)), so I couldn't use `length' on it. That and the
fact that the strings have a face property made it seem like the
function wasn't really meant for what I was doing with it

>> but they're absolutely useful in a programmatic setting. I wonder if
>> there could be some version of these functions that could be "blessed"
>> for use in Elisp programs.
>
> They're definitely allowed to be used for ELisp programs.  They're not
> 100% pure functions, but not too far off.  The main issue I can see for
> use "internal use" is that their behavior is influenced by the
> `completion-styles` user-config, which may or may not be what you want.

I'll wait and see if that's a problem. If I'm expecting a completion
style that completes hyphenated strings ("con-f" -> "contact-from") and
the user has removed that style it could be an issue. I wonder how many
users even know about this configuration, though.

>>> (defvar gnus-search-minibuffer-map
>>>       (let ((km (make-sparse-keymap)))
>>>         (set-keymap-parent km minibuffer-local-map)
>>>         (define-key km (kbd "SPC") #'self-insert-command) ;; Isn't this 
>>> redundant?
>>
>> Somewhere I'd gotten the idea that SPC was bound to
>> `minibuffer-complete-word', I don't know how.
>
> I presume you earlier inherited from `minibuffer-local-completion-map`
> or something like that.

Yes, I was confusing with `completing-read'.

>> Ah, of course! And if `gnus-search--complete-key-data' is conservative
>> about when it fires, it leaves the door open for other completion
>> functions.
>
> If by "fires" you mean "returns non-nil", then yes, indeed.
>
>> This function could complete "cont" to "contact:", at which
>> point (for example) an EBDB-specific capf function could take over and
>> complete names or email addresses to search for.
>
> Right.  Or it could itself recognize "contact:" and return the bounds of
> the contact info along with EBDB's completion table (since EBDB's
> capf presumably doesn't know about the "contact:" syntax).

Right, it would be much easier just to have a variable where
contact-management packages can register a completion table. Now I'm
thinking there's no real need for a "contact" key at all, just offer
completion on to/from/recipient/etc. Anyway, I've already got the code
working, thanks again for the pointers.

Eric