Re: Code for cond*

[João Távora]

On Thu, Jan 18, 2024 at 3:44 PM Andrea Corallo <acorallo@gnu.org> wrote:

> apologies if it was discussed already, wanted to ask: what is the reason
> for some of these cond* clauses to keep the binding in effect outside
> the clause itself and for the whole cond* construct?  At first glance it
> doesn't look very idiomatic in Lisp terms to me.

I think I asked the same, and Drew Adams and Adam Porter also
raised this point:

   https://lists.gnu.org/archive/html/emacs-devel/2023-12/msg00554.html
   https://lists.gnu.org/archive/html/emacs-devel/2023-12/msg00618.html

and I'm not sure we eventually clarified it.  Apologies if I missed
some email addressing it as I'm not following this discussion in
detail.

Also, since cond* does a strict subset of what pcase does, has any
thought been given to implementing 'cond*' _on top_ of 'pcase', like
some systems define 'cond' in terms of 'if' (or vice versa)?

Not meant as gratuitous provocation, really.  I think it'd have advantages:

* presumably every cond* feature can be supported (if indeed I'm right
  about the "strict subset")

* it would probably save a lot of repeated code

* it would make it easy for those who can already read pcase
  to just macroexpand the cond* form and see what it is actually
  doing underneath.

So could actually facilitate the adoption path for 'cond*' (as
questionable as that path may still be, at least for some parties).

João