GNU bug report logs - #51655
[PATCH] build-system: haskell: Add ‘package-with-explicit-haskell’ procedure.

Previous Next

Package: guix-patches;

Reported by: Xinglu Chen <public <at> yoctocell.xyz>

Date: Sun, 7 Nov 2021 10:56:01 UTC

Severity: normal

Tags: patch

Full log


Message #14 received at 51655 <at> debbugs.gnu.org (full text, mbox):

From: Xinglu Chen <public <at> yoctocell.xyz>
To: zimoun <zimon.toutoune <at> gmail.com>, 51655 <at> debbugs.gnu.org
Cc: Lars-Dominik Braun <lars <at> 6xq.net>
Subject: Re: [bug#51655] [PATCH] build-system: haskell: Add
 ‘package-with-explicit-haskell’ procedure.
Date: Sun, 19 Dec 2021 09:47:22 +0100
[Message part 1 (text/plain, inline)]
Hi,

On Mo, Nov 15 2021, zimoun wrote:

> Hi,
>
> On Sun, 07 Nov 2021 at 11:55, Xinglu Chen <public <at> yoctocell.xyz> wrote:
>> * guix/packages.scm (package-with-explicit-compiler): New procedure;
>> factorized from ‘package-with-explicit-python’.
>> * guix/build-system/python.scm (package-with-explicit-python): Use
>> ‘package-with-explicit-compiler’.
>> * guix/build-system/haskell.scm (package-with-explicit-haskell): New procedure.
>> * doc/guix.texi (Defining Package Variants): Document it.
>> ---
>>  doc/guix.texi                 | 37 ++++++++++++++++++
>>  guix/build-system/haskell.scm |  8 ++++
>>  guix/build-system/python.scm  | 71 +++++++----------------------------
>>  guix/packages.scm             | 55 +++++++++++++++++++++++++++
>>  4 files changed, 113 insertions(+), 58 deletions(-)
>
> Neat refactoring! :-)  It is something discussing several times and I do
> not know if ’package-with-explicit-ocaml’ could not be added too.  The
> issue is that it depends on more than only one element for “compiling”:
> ocaml, findlib and dune.

Ah, yes, I forgot about OCaml.  :-)

> On the other hand, guix/build-system/gnu.scm contains
> ’package-with-explicit-inputs*’ which is more or less the same idea.
> And ’transform-package-toolchain’ is another instance.

I didn’t about ‘package-with-explicit-inputs*’, but I will look into it. 

> This ’package-with-explicit-{ocaml,python}’ pre-dates the
> transformations and deep rewriting.
>
> Therefore, maybe refactor could revisit the complete story.
>
> Well, I had headaches last time I gave a look at transformations.  And I
> do not know if it is possible to have a generic transformation for
> rewriting implicit inputs of all build systems.  That’s something that
> could be nice to nice: ’package-with-explicit-python’ based on a
> “generic” ’package-with-explicit-compiler’ (compiler or toolchain
> though) and provides ’package-with-explicit-*’ for all build systems
> (ocaml, ghc, emacs, guile, etc.).

Yeah, that would be nice to have.  I currently have some other things I
would like to focus on, but I should look more into package
transformations, and hopefully come up with a good solution to this.
:-)

>> +@lisp
>> +(use-modules (guix build-system haskell))
>> +
>> +(define (default-haskell-8.8)
>> +  ;; Lazily resolve the binding to avoid a circular dependency.
>> +  (let ((haskell (resolve-interface '(gnu packages haskell))))
>> +    (module-ref haskell 'ghc-8.8)))
>> +
>> +(define package-with-haskell-8.8
>> +  (package-with-explicit-haskell (delay (default-haskell-8.8))
>> +                                 "ghc-" "ghc-8.8-"
>> +                                 #:variant-property 'ghc-8.8-variant))
>> +@end lisp
>
> As Lars said, an instance is missing, IMHO.

Not sure what you are referring to here, what exactly do you mean with
“instance”?
[signature.asc (application/pgp-signature, inline)]

This bug report was last modified 3 years and 178 days ago.

Previous Next


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