On Thu, 15 Feb 2018 23:21:42 -0500,
Eli Schwartz via arch-projects wrote:
>
> This looks reasonable, thanks. BTW no need to send a cover letter for
> one patch. :)
I felt silly sending it, but I wanted to note that your fixes looked
good to me for this, but I didn't want to put that in the commit
message. :)
> > ---
> > test/cases/db-update.bats | 11 ++-
> > 1 file changed, 10 insertions(+), 1 deletion(-)
> >
> > diff --git a/test/cases/db-update.bats b/test/cases/db-update.bats
> > index 1da7eef..6604841 100644
> > --- a/test/cases/db-update.bats
> > +++ b/test/cases/db-update.bats
> > @@ -92,7 +92,16 @@ load ../lib/common
> > db-update
> > checkPackage extra pkg-any-a
> >
> > - releasePackage extra pkg-any-a
> > + # don't let __buildPackage use the cached build; we want to
> > + # force a new build with a different PKGEXT.
> > + if [[ -n ${BUILDDIR} ]]; then
> > + mv -T "${BUILDDIR}/$(__getCheckSum
> > "${TMP}/svn-packages-copy/pkg-any-a/trunk/PKGBUILD")"{,.bak}
> > + fi
> > + PKGEXT=.pkg.tar.gz releasePackage extra pkg-any-a
> > + if [[ -n ${BUILDDIR} ]]; then
> > + rm -rf "${BUILDDIR}/$(__getCheckSum
> > "${TMP}/svn-packages-copy/pkg-any-a/trunk/PKGBUILD")"
> > + mv -T "${BUILDDIR}/$(__getCheckSum
> > "${TMP}/svn-packages-copy/pkg-any-a/trunk/PKGBUILD")"{.bak,}
> > + fi
> > run db-update
> > [ "$status" -ne 0 ]
> > }
>
> I'm guessing you restore the old version because the new version will
> throw off other tests using the cached version with the wrong $PKGEXT?
> It might make more sense to change __buildPackage() to use is_globfile
> ${cache}*${PKGEXT} rather than [[ -d ${cache} ]]
>
> This should prevent needing to move anything, leading to cleaner code on
> both sides.
That's a good idea. I'll do that!
--
Happy hacking,
~ Luke Shumaker
