On Mon, May 13, 2019 at 01:07:07PM -0400, Kurt Mosiejczuk wrote:
> During our work on the portgen improvements afresh1 mentioned that the
> cpan module just adds RUN_DEPENDS to TEST_DEPENDS automatically. So I
> coded up this patch to lang/python/python.port.mk to do the same.
>
> It will still obey MODPY_TESTDEP being set to no, in which case we
> don't touch TEST_DEPENDS.
>
> Thoughts? If folks like this, I can go through and update python port
> Makefiles that do this manually right now and commit this change and the
> Makefile changes at one time.
As a heavy user of cpan.port.mk I was surprised this didn't exist more
generally, it seems like a good idea, but I don't understand why
MODPY_TEST_DEPENDS exists and such, but I do know it's very common to
need this.
$ grep '^TEST_DEPENDS.*=.*RUN_DEPENDS' */py-*/Makefile | wc -l
116
Although I don't know what makes it required, but it's possible it's
MODPY_PYTEST.
$ ls */py-*/Makefile | wc -l
652
>
> --Kurt
>
> Index: lang/python/python.port.mk
> ===================================================================
> RCS file: /cvs/ports/lang/python/python.port.mk,v
> retrieving revision 1.108
> diff -u -p -r1.108 python.port.mk
> --- lang/python/python.port.mk 27 Apr 2019 21:28:30 -0000 1.108
> +++ lang/python/python.port.mk 13 May 2019 16:31:21 -0000
> @@ -72,10 +72,9 @@ MODPY_RUN_DEPENDS = lang/python/${MODPY_
> MODPY_LIB_DEPENDS = lang/python/${MODPY_VERSION}
> _MODPY_BUILD_DEPENDS = lang/python/${MODPY_VERSION}
>
> +MODPY_TEST_DEPENDS = ${RUN_DEPENDS}
> .if ${MODPY_PYTEST:L} == "yes"
> -MODPY_TEST_DEPENDS = devel/py-test${MODPY_FLAVOR}
> -.else
> -MODPY_TEST_DEPENDS = # Leave blank for now for non-pytest
> +MODPY_TEST_DEPENDS += devel/py-test${MODPY_FLAVOR}
> .endif
>
> .if ${NO_BUILD:L} == "no"
>
--
andrew - http://afresh1.com
Software doesn't do what you want it to do, it does what you tell it do.
-- Stefan G. Weichinger.
