On Thu, 28 Jun 2018 15:27:04 +0100, Stuart Henderson
<[email protected]> wrote:
> > > > This has a conflict:
> > > >
> > > > $ make plist
> > > > Warning: py-magic-0.4.15 conflicts with py-libmagic-5.32
> > > > (devel/py-libmagic):/usr/local/lib/python2.7/site-packages/magic.py
> > > > /usr/local/lib/python2.7/site-packages/magic.pyc
> > > > /usr/local/lib/python2.7/site-packages/magic.pyc
> > > > /usr/ports/openbsd-wip/devel/py-magic/pkg/PLIST
> > > > changed
> > >
> > > Yes and py-libmagic is only needed by print/py-relatorio so my
> > > plan was to check if py-relatorio could be switched to it but I
> > > didn't go further than asking semarie about it.
> >
> > there is no problem for py-relatorio for switching to py-magic
> > instead of py-libmagic. And in fact, py-magic is the intented
> > dependency for upstream (and from reading py-relatorio code, it is
> > possible that with py-libmagic, it doesn't work as expected).
> >
> > I ran the testsuite of py3-relatorio with both (py3-magic and
> > py3-libmagic), and the result are same: all 23 tests passes.
> >
> > If py-libmagic has no other reverse dependencies than py-relatorio,
> > I would be in favor to remove it, and use py-magic as dependency
> > instead of. Below a diff that do it, and while here taking
> > maintainership of py-relatorio.
>
> Great, I'm OK with that approach.
>
> I think if we're doing this, it would make sense to auto update
> py-libmagic to py-magic. We'll need some magic in devel/py-magic so
> that conflict/pkgpath lines are set correctly, I think this should do
> the trick:
>
> @conflict ${MODPY_PY_PREFIX}libmagic-*
> @pkgpath devel/py-libmagic${MODPY_FLAVOR}
I'm ok as well with this move and the new port.
