Re: [libvirt PATCH 11/33] ci: build.sh: Add a wrapper function over the 'rpmbuild' job

On Thu, Aug 31, 2023 at 05:55:54PM +0100, Daniel P. Berrangé wrote: > On Fri, Aug 25, 2023 at 07:55:19PM +0200, Erik Skultety wrote: > > This helper is a shell function transcript of its original GitLab CI > > counterpart. > > > > Signed-off-by: Erik Skultety <eskultet(a)redhat.com&gt; > > --- > > ci/build.sh | 11 +++++++++++ > > 1 file changed, 11 insertions(+) > > > > diff --git a/ci/build.sh b/ci/build.sh > > index 6990f2d171..30f4712e4b 100644 > > --- a/ci/build.sh > > +++ b/ci/build.sh > > @@ -94,3 +94,14 @@ run_potfile() { > > run_meson_setup > > run_build > > } > > + > > +run_rpmbuild() { > > + local CMD="rpmbuild \ > > + --clean \ > > + --nodeps \ > > + --define "_without_mingw 1" \ > > + -ta build/meson-dist/libvirt-*.tar.xz" > > + > > + run_meson_setup > > Redundant here as implied by run_dist While I can drop all these redundant setup calls, the reason why I put them there was simply readability. Let's face it Shell isn't a particularly nice structured language to look at especially when it comes to functions, so rather than having everyone go and look what run_dist/run_build, etc. do, I instead opted for transparent naming and explicit function calls, IOW so that when you look at run_rpmbuild you immediately know we run meson setup followed by a project build, etc., but then if you call run_dist alone in a local interactive environment then run_dist would also call meson_setup because it can't go without it, so hence the redundancy in all the functions. If you still feel like it's undesirable even in ^this case, I will drop the calls.