bug#40693: 28.0.50; json-encode-alist changes alist

[Dmitry Gutov]

On 29.04.2020 15:21, Eli Zaretskii wrote:
> > Cc: darthandrus@gmail.com, 40693@debbugs.gnu.org
> > From: Dmitry Gutov <dgutov@yandex.ru>
> > Date: Wed, 29 Apr 2020 15:08:57 +0300
> >
> > > No one tried to come up with arguments why this has to be in emacs-27.
> >
> > Let me try:
> >
> > It fixes a bug, one which could be annoying to investigate, the fix is
> > small and localized to the case when json-encoding-object-sort-predicate
> > is non-nil (so pretty safe).
>
> It also makes the function slower.  Which may be an important issue
> for JSON processing.  Callers that don't care about the original list
> will be "punished" regardless.

It's a somewhat fair point (copy-sequence is much faster than sort, 
usually, but if we include GC time, it can be significant). The way 
json-encoding-object-sort-predicate is implemented, though, it's not 
very performance-oriented.

Applications like lsp-mode aren't going to use this variable, so they 
shouldn't be hit (and it isn't supported in native JSON either).

On that subject, users really shouldn't set it either. Rather, any 
user-facing feature that outputs JSON, when this behavior could be 
desirable, should have its own tweaking knob.

> How about adding an optional argument instead, by default off, to
> request this behavior?  then callers who care about the original alist
> could request a non-destructive operation, and others won't suffer any
> slowdown.

The current behavior is unsafe, that's the problem. Also, 
json-encode-alist is called in a recursive fashion, so it'd have to be a 
global variable instead.