[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
bug#11938: 24.1.50; Some not so useful CL related byte compiler warnings
From: |
Stefan Monnier |
Subject: |
bug#11938: 24.1.50; Some not so useful CL related byte compiler warnings |
Date: |
Wed, 18 Jul 2012 06:25:16 -0400 |
User-agent: |
Gnus/5.13 (Gnus v5.13) Emacs/24.1.50 (gnu/linux) |
> If I byte compile it, I get the following warnings:
> progv-bug.el:7:37:Warning: global/dynamic var `body' lacks a prefix
> progv-bug.el:10:25:Warning: reference to free variable `a'
> progv-bug.el:10:27:Warning: reference to free variable `b'
> progv-bug.el:11:35:Warning: reference to free variable `year'
> progv-bug.el:19:1:Warning: `labels' is an obsolete function (as of 24.2); use
> `(quote cl-labels)' instead.
Yay!! So there are progv users after all! Amazing!
> In detail:
> progv-bug.el:7:37:Warning: global/dynamic var `body' lacks a prefix
> Dunno what that means, the code doesn't include such a variable.
Yes, it's a bug of my new progv code, which introduces this var.
Should be fixed now, thank you.
> progv-bug.el:10:25:Warning: reference to free variable `a'
> progv-bug.el:10:27:Warning: reference to free variable `b'
> progv-bug.el:11:35:Warning: reference to free variable `year'
> Not very useful, since the variables' values are never looked up
> before they are bound, but ok, I can live with it.
Yes, I think these warnings were already present with Emacs-24.1's
version of progv.
I'm curious, actually. What does Common-Lisp do with something like:
(let ((a 2)) (progv '(a b) '(1 2) a))
does it return 2 (the value of the lexically-scoped var `a') or 1?
If it should return 1, what about:
(let ((a 2)) (progv '(b) '(1) a))
In any case, for all uses of progv where the first arg is a constant,
you'd be much better off using pcase-let (or destructuring-bind or
multiple-value-bind).
So maybe you're not a user of progv after all.
> progv-bug.el:19:1:Warning: `labels' is an obsolete function (as of 24.2);
> use
> `(quote cl-labels)' instead.
> What does `(quote cl-labels)' mean? I guess this should be just
> `cl-labels'.
Indeed, that was a copy&paste typo, thank you for catching it,
Stefan