Re: [PATCH 3/3] meson: don't install git-pack-redundant(1) docs with breaking changes
- From
Patrick Steinhardt <ps@pks.im>
- Date
- Mar 19, 2025, 09:37 UTC
- Message-ID
- <Z9qQQAhFk-MNwiom@pks.im>
- In-Reply-To
- <7fee83f7-995b-4c84-9216-caa6803a69d9@gmail.com>
On Tue, Mar 18, 2025 at 10:06:51AM +0000, Phillip Wood wrote:
Show 23 quoted lines
> On 17/03/2025 14:50, Phillip Wood wrote:
> >
> > Oh so if we have selected breaking_changes then manpages.keys() does not
> > include "git-pack-redundant.adoc" but that file exists and so we need to
> > add it to the list of configured man pages. If breaking_changes is
> > selected then don't we end up adding "git-pack-redundant.adoc" to
> > configured_manpages twice? Does that matter when we come to do
> >
> > actual_manpages = run_command(shell, '-c', 'ls git*.adoc scalar.adoc',
> > check: true,
> > env: script_environment,
> > ).stdout().strip().split('\n')
> >
> > if configured_manpages != actual_manpages
> > ...
> >
> > ? Also I'm confused as to how that comparison works without sorting
> > configured_manpages. Even if manpages.keys() sorts the keys (the
> > documentation at [1] is silent on that) we add some out-of-order entries
> > to the end of the list.
>
> I think the answer is that the comparison always fails but as there are no
> missing or superfluous man pages the body of the if does not error out.Yeah. We indeed may have it multiple times now, but as you noticed it ultimately still works. I didn't care too deeply to avoid the duplication because in the end this step is only used to verify that we have all manpages wired up in Meson.
Patrick