logilab-common #269904 new test failures under release logilab-common-0.62.1 [waiting feedback]

Comments

• 2014/11/03 09:41, written by jcristau

  please try current head. the unittest_date error should be fixed by http://hg.logilab.org/master/logilab/common/rev/8944b8da95a8 and I'm pretty sure the others pass too, at least on 2.7 and 3.4.
  • 2014/11/07 03:45, written by idella4

    Is this head?
    

    hg clone http://hg.logilab.org/review/logilab/common/
    

    What ever it is it yields, under py3.4;
    

    *******************************************************************************
    Ran 372 test cases in 12.02s (0.34s CPU), 1 errors, 4 failures, 1 skipped
    19 modules OK (3 failed)
    failures: /home/testuser/github/common/test/unittest_umessage [3/4],
    /home/testuser/github/common/test/unittest_date [1/1],
    /home/testuser/github/common/test/unittest_configuration [1/17]
    

    I rebased the patch of 8944b8da95a8, added it to the logilab-common-0.62.1.ebuild and it proved ineffective.
    Yielded similar to the above. I think this is addressing a different issue,
    The tests of unittest_configuration unittest_date ofc. match my initial report so I'm guessing this isn't head.
    

    However, the good news. What logilab.org/review/logilab/common/ does appear to have cured is the one thing I didn't
    actually add in the report; issue 241813 by Arfrever. However I am still confused. The issue pertains to the class MxDateTC in unittest_date.py. In that issue the The "raise SkipTest(reason)" was cited. Here is one of the pair of errors I get from the run under py3.4;
    

    ========================= unittest_date.py =========================
    .......................unhandled exception occurred while testing unittest_date
    Traceback (most recent call last):
    File "/usr/lib64/python3.4/site-packages/logilab/common/testlib.py", line 550, in quiet_run
    func(*args, **kwargs)
    File "/mnt/gen2/TmpDir/portage/dev-python/logilab-common-0.62.1/work/logilab-common-0.62.1-python3_4/build/test/lib/logilab_common-0.62.1-py3.4.egg/logilab/common/test/unittest_date.py", line 160, in setUp
    self.check_mx()
    File "/mnt/gen2/TmpDir/portage/dev-python/logilab-common-0.62.1/work/logilab-common-0.62.1-python3_4/build/test/lib/logilab_common-0.62.1-py3.4.egg/logilab/common/test/unittest_date.py", line 157, in check_mx
    self.skipTest('mx.DateTime is not installed')
    File "/usr/lib64/python3.4/unittest/case.py", line 638, in skipTest
    raise SkipTest(reason)
    unittest.case.SkipTest: mx.DateTime is not installed
    

    During handling of the above exception, another exception occurred:
    

    Traceback (most recent call last):
    File "/usr/lib64/python3.4/site-packages/logilab/common/pytest.py", line 1134, in _ts_wrapped_run
    test(result, runcondition, options)
    File "/usr/lib64/python3.4/site-packages/logilab/common/testlib.py", line 586, in __call__
    if not self.quiet_run(result, self.setUp):
    File "/usr/lib64/python3.4/site-packages/logilab/common/testlib.py", line 554, in quiet_run
    self._addSkip(result, str(e))
    TypeError: _addSkip() missing 1 required positional argument: 'reason'
    

    During handling of the above exception, another exception occurred:
    

    This
    TypeError: _addSkip() missing 1 required positional argument: 'reason'
    

    I thought was the essence of issue 241813, now I am just confused. There is an obvious mis-match between
    py3.3 amd 3.4 in the number of args in the code.
    Difference between unittest/case.py in Python 3.3 and 3.4 contains:
    

    - def _addSkip(self, result, reason):
    + def _addSkip(self, result, test_case, reason):
    (from issue 241813)
    

    All I know is this still carries over in logilab-common-0.62.1 in 3 of the 4 of unittest_umessage, py2.7 has
    the 1 Fail under unittest_configuration.py and py3.4 is a mess.
    

    I normally don't clone repos, so if you're inclined to supply me with a 'correct' repo if I''ve used
    used a repo other than head, feel free to offer it here.