GNU bug report logs - #49181
Fix missing phases in Emacs builds

Previous Next

Package: guix-patches;

Reported by: Carlo Zancanaro <carlo <at> zancanaro.id.au>

Date: Wed, 23 Jun 2021 06:46:01 UTC

Severity: normal

Done: Maxim Cournoyer <maxim.cournoyer <at> gmail.com>

Bug is archived. No further changes may be made.

Full log


Message #14 received at 49181 <at> debbugs.gnu.org (full text, mbox):

From: Sarah Morgensen <iskarian <at> mgsn.dev>
To: Leo Prikler <leo.prikler <at> student.tugraz.at>
Cc: 49181 <at> debbugs.gnu.org, carlo <at> zancanaro.id.au
Subject: Re: bug#49181: Fix missing phases in Emacs builds
Date: Fri, 23 Jul 2021 14:32:08 -0700
Hi!

I'm glad to see this.  This behavior has caught me up a number of times
recently.

Leo Prikler <leo.prikler <at> student.tugraz.at> writes:

[...]

 @@ -571,18 +572,22 @@ effects, such as displaying warnings or error messages."
>  (define* (alist-cons-before reference key value alist
>                              #:optional (key=? equal?))
>    "Insert the KEY/VALUE pair before the first occurrence of a pair whose key
> -is REFERENCE in ALIST.  Use KEY=? to compare keys."
> +is REFERENCE in ALIST.  Use KEY=? to compare keys.  An error is raised when no
> +such pair exists."
>    (let-values (((before after)
>                  (break (match-lambda
>                          ((k . _)
>                           (key=? k reference)))
>                         alist)))
> -    (append before (alist-cons key value after))))
> +    (match after
> +      ((reference after ...)
> +       (append before (alist-cons key value (cons reference after)))))))

This can probably just be (untested):

      ((reference after* ...)
       (append before (alist-cons key value after))))))

Or if we want to avoid extraneous bindings completely (also untested):

      ((_ _ ...)
       (append before (alist-cons key value after))))))

>  
>  (define* (alist-cons-after reference key value alist
>                             #:optional (key=? equal?))
>    "Insert the KEY/VALUE pair after the first occurrence of a pair whose key
> -is REFERENCE in ALIST.  Use KEY=? to compare keys."
> +is REFERENCE in ALIST.  Use KEY=? to compare keys.  An error is raised when
> +no such pair exists."
>    (let-values (((before after)
>                  (break (match-lambda
>                          ((k . _)
> @@ -590,9 +595,7 @@ is REFERENCE in ALIST.  Use KEY=? to compare keys."
>                         alist)))
>      (match after
>        ((reference after ...)
> -       (append before (cons* reference `(,key . ,value) after)))
> -      (()
> -       (append before `((,key . ,value)))))))
> +       (append before (cons* reference `(,key . ,value) after))))))
>  
>  (define* (alist-replace key value alist #:optional (key=? equal?))
>    "Replace the first pair in ALIST whose car is KEY with the KEY/VALUE pair.

Other than that it looks good to me. Pushing this should also close #32661.

This should be a patch for core-updates, though, since it changes
derivations for any package that (directly or indirectly) uses
ALIST-CONS-BEFORE or ALIST-CONS-AFTER.

--
Sarah




This bug report was last modified 1 year and 229 days ago.

Previous Next


GNU bug tracking system
Copyright (C) 1999 Darren O. Benham, 1997,2003 nCipher Corporation Ltd, 1994-97 Ian Jackson.