Re: Adding python3-biopython
Hi Peter,
thanks for your answer which finally is right in time since I'm
currently in the last phase before I would like to upload the BioPython
packages with Python 3.3 and 3.4 support as well as running an as
complete as possible test suite.
On Sun, Apr 27, 2014 at 01:21:14PM +0100, Peter Cock wrote:
> On Wed, Mar 12, 2014 at 10:23 AM, Andreas Tille <andreas@an3as.eu> wrote:
> > Hi again Peter,
> >
> > I have again detected an issue in the test suite which seems to be
> > caused by a missing file:
> >
> > test_Fasttree_tool
> > ----------------------------------------------------------------------
> > Traceback (most recent call last):
> > File "run_tests.py", line 301, in runTest
> > self.fail("Warning: Can't open %s for test %s" % (outputfile, self.name))
> > AssertionError: Warning: Can't open ./output/test_Fasttree_tool for test test_Fasttree_tool
> >
> >
> > The file Tests/output/test_Fasttree_tool is missing from your source
> > tarball (as well from the Git repository). Since we really like to
> > create the best possible package we would really try to run the full
> > test suite and so it would be great if you somehow could provide this
> > file (or give a hint how we could otherwise successfully run the test.
>
> Correct. I forget the circumstances, but this was fixed last month:
> https://github.com/biopython/biopython/commit/2b414d8a9ca0aa15ef4b4b3680b8bfc755776b56
Thanks for the hint. I injected this into the Debian packaging.
However, the Fasttree test keeps on failing - now for a different
reason:
======================================================================
FAIL: runTest (__main__.ComparisonTestCase)
test_Fasttree_tool
----------------------------------------------------------------------
Traceback (most recent call last):
File "run_tests.py", line 321, in runTest
assert expected_line == '', "Unread: %s" % expected_line
AssertionError: Unread:
----------------------------------------------------------------------
I admit I'm not very familiar with Python test suites. I'd volunteer to
do some more detailed debugging but since print does not work in test
suites I have no idea how to issue some debugging status lines to find
the exact reason for the failure.
> > I also repeat my question from my last mail:
> >
> > Is it OK if I send this kind of reports to you or should I rather use
> > the bug tracker or any mailing list. Sorry, I have not maintained the
> > package before and thus do not know the prefered way of communication.
>
> Personal email is fine, but as in this thread I do sometimes miss things.
> You can report issues via GitHub, or raise queries on the Biopython
> mailing list:
>
> https://github.com/biopython/biopython/issues
> http://lists.open-bio.org/mailman/listinfo/biopython/
I added these to Upstream-Contact metadata field to make sure I or other
people who might touch the Debian package will remember in the future.
I have an additional question about the BioPython package layout where I
would highly evaluate your input. Today I sended the following question
to the Debian Med mailing list:
https://lists.debian.org/debian-med/2014/04/msg00094.html
What is your opinion: Should all these tools which are used in the test
suite be installed on a "typical" (whatever typical means in your eyes)
BioPython installation?
Kind regards
Andreas.
--
http://fam-tille.de
Reply to: