bug#58950: [PATCH] * lisp/subr.el (buffer-match-p): Optimise performance

[Dmitry Gutov <dgutov@yandex.ru>]

On 31/12/2022 15:56, Philip Kaludercic wrote:
> Dmitry Gutov<dgutov@yandex.ru>  writes:
> 
>> On 01.11.2022 21:11, Philip Kaludercic wrote:
>>> 1. Style.  I wrap the defun in a let (or rather letrec) block to avoid
>>>      littering the global namespace.  It isn't necessary, and one could
>>>      argue it makes debugging more difficult.
>>> 2. Caching policy.  Caching is critical to this optimisation.  Just
>>>      using byte-compilation would cause the above test to slow down to
>>>      (76.323692627 656 57.088315405).  The question is if the hash map
>>>      will collect too much garbage over time, and if there is a better
>>>      approach that could be taken?
>> I'd like to let our language-level specialists to take the deeper look.
>>
>> This approach looks the most straightforward, but there could be
>> others, just like "compiling" the form inside defcustom setter (for
>> project-kill-buffer-conditions, and every similar option), doing
>> precompilation closer to where the rules are used (similar to
>> font-lock-compile-keywords), or not doing any of that. All depending
>> on how long a typical compilation takes, and how many buffers the user
>> has to have, to see any noticeable benefit.
>>
>> On the last note, I'm curious how many buffers would it take to see a
>> 50ms improvement in match-buffers' runtime when using the current
>> project-kill-buffer-conditions's value, for example.
> Ping?  If this change is too controversial, I'd like to backport the
> changes from bug#58951 and apply them since they are fixing an actual bug.

It does look too ambitious for the release branch.

On the merits of the change itself, maybe Mattias or Stefan have some 
thoughts?