--- Begin Message ---
Subject: |
false-positive warning "variable ‘_’ not left unused" in if-let* and if-let |
Date: |
Wed, 14 Feb 2024 00:21:01 +0300 |
User-agent: |
Evolution 3.50.3 |
I've been writing an answer for a question on emacs.stackexchange¹ and to avoid
nested `if` and `let` clauses I used a `if-let*`, and result of one of the
checks I
assigned to a `_` variable, because the variable would be left unused, it's
only the
check being non-nil that mattered.
But when byte-compiled that triggered a:
test.el:6:9: Warning: variable ‘_’ not left unused
…which is untrue, because it is unused.
The problem is present in both `if-let` and `if-let*`
# Steps to reproduce
1. Create test.el with the following code:
;;; -*- lexical-binding: t -*-
(if-let*
((_ nil))
(print "then clause")
(print "else clause"))
2. M-x byte-compile test.el
## Expected
It byte-compiles with no warnings
## Actual
It compiles with a warning:
test.el:3:7: Warning: variable ‘_’ not left unused
# Additional information
Emacs version: commit d4d5830f8a0 built two weeks ago from master.
1: https://emacs.stackexchange.com/questions/80351/delete-prettify-symbol
--- End Message ---
--- Begin Message ---
Subject: |
Re: bug#69108: false-positive warning "variable ‘_’ not left unused" in if-let* and if-let |
Date: |
Sun, 25 Feb 2024 09:42:07 +0200 |
> From: Michael Heerdegen <michael_heerdegen@web.de>
> Cc: 69108@debbugs.gnu.org, Hi-Angel@yandex.ru
> Date: Sun, 25 Feb 2024 02:54:04 +0100
>
> Eli Zaretskii <eliz@gnu.org> writes:
>
> > > [...] repeating a test. Is this what you prefer?
> >
> > Yes, I think so. And you could perhaps avoid repetition of
> > (cdr binding) by saving the result of (null (cdr binding)) in
> > a local variable.
>
> I went with a separate `cond' branch added instead, I think this is even
> simpler. Ok?
Yes, thanks.
> > > {My unsuccessful doc tweaks}
> >
> > [...] I'd prefer to separate it.
>
> Done. Feel free to tune it to your likes. Or send me an "ok", then
> I'll just commit this version.
I installed them both, thanks.
Closing the bug.
--- End Message ---