[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: this-package-input vs. propagated-inputs in package definition
From: |
Simon Tournier |
Subject: |
Re: this-package-input vs. propagated-inputs in package definition |
Date: |
Tue, 10 Sep 2024 19:14:44 +0200 |
Hi,
On Mon, 09 Sep 2024 at 11:27, Andy Tai <atai@atai.org> wrote:
> A question on package definition (likely scope of guix-devel, but
> anyway posted here first):
>
> I have updated the definition of package ddd which has libxft as
> input. Libxft in turn has freefont as propagated-input. In the
> definition of ddd I tried to do
>
> (this-package-input "freefont")
>
> which returns #f.
I think that’s expected. ’this-package-input’ searches only in the list
of ’inputs’ and ’propagated-inputs’ of the package itself and not of the
inputs. Somehow, it’s not recursive and “freetype” is not an “inputs”
of the package ddd.
BTW, I am not sure to understand what mean a propagated-inputs in the
context of a build. I mean, it sounds at run-time, not build-time. No?
Do I miss something?
> I had to add freefont as input to ddd to make this work.
BTW, for the package ddd, is freetype an input or also a
propagated-input ?
> As propagated-input should be propagated, shall the addition of
> freefont as input to ddd be unneeded?
Well, for sure, considering the current implementation, you cannot reach
the propagated-inputs of the inputs using this-package-inputs.
Cheers,
simon