[Zope-dev] Re: Broken Unittest

[Gary Poster]

That sounds possible.  I added support of Python 2.3 datetime, and while 
I tried to handle alternate timezones correctly I may have failed.  I'll 
try to look at this later today.

Gary

Tres Seaver wrote:
> Christian Theune wrote:
> 
>> Hi,
>>
>> I address you directly because I haven't received feedback on this from
>> the zope-dev mailinglist and you have been the last ones touching code
>> in Products/DateIndex.
>>
>> This test is breaking for me but not on Jeremys daily tests for Zope 2
>> on the HEAD:
>>
>> ======================================================================
>> FAIL: test_retrieval
>> (Products.PluginIndexes.DateIndex.tests.test_DateIndex.DI_Tests)
>> ----------------------------------------------------------------------
>> Traceback (most recent call last):
>>   File
>> "/home/ctheune/Development/Zope/Zope/lib/python/Products/PluginIndexes/DateIndex/tests/test_DateIndex.py", 
>> line 184, in test_retrieval
>>     self.failUnlessEqual(len(index), len(values) - 2) # One dupe, one
>> empty
>>   File "/usr/lib/python2.3/unittest.py", line 302, in failUnlessEqual
>>     raise self.failureException, \
>> AssertionError: 8 != 9
>>
>> ----------------------------------------------------------------------
>>
>> I tried debugging that but I don't know the internal interfaces of the
>> PluginIndexes well enough to not waste time.
>>
>> I don't know which special information you could need. I'm running
>> pretty everything standard (python 2.3, linux with 2.6.3 kernel).
>>
>> The Sandbox was compiled by:
>>
>>   ./configure
>>   make
>>   make instance
>>
>> The test was run by
>>
>>   python2.3 test.py
>>
>> This test is the only one that breaks and keeps me from checkin in
>> several changes for quite a while already.
> 
> 
> That test isn't broken for me on the HEAD.  At a guess, I would bet that 
> some change (re)introduced a sensitivity to timezones within the test; 
> can you try running the test in an environment where the timezone is 
> GMT?   GMT-1?  I think you are in GMT+1 land, right?
> 
> Tres.