Lieven Govaerts wrote:
> Vlad Georgescu wrote:
>> Lieven Govaerts wrote:
>>
>>> vgeorgescu@tigris.org wrote:
>>>
>>>> + # Update at depth=immediates in the other wc, expecting to see no errors.
>>>> + svntest.actions.run_and_verify_svn("Output on stderr where none expected",
>>>> + SVNAnyOutput, [],
>>>> + 'update', '--depth', 'immediates',
>>>> + other_wc)
>>>> +
>>>>
>>> Vlad,
>>>
>>> I think this test would be (even) better if you actually checked the
>>> output of update, it's supposed to receive the property changes on A right?
>>>
>> Yes, but the last update fails because of an error:
>>
>> Output on stderr where none expected
>> EXPECTED STDERR:
>> ACTUAL STDERR:
>> ../svn-trunk/subversion/libsvn_wc/lock.c:989: (apr_err=155005)
>> svn: Unable to lock 'svn-test-work/working_copies/depth_tests-18.other/A'
>> EXCEPTION: SVNLineUnequal
>>
>> If I change the test to use expected_output, will it still display
>> the error, or will it just complain about the trees being unequal?
>>
> Hm, had to check that. In svntest/actions.py the run_and_verify_update
> function first matches the error and only then starts comparing the trees.
If I'm reading the code right, if the error matches error_re_string,
run_and_verify_update returns and doesn't compare the trees (which means
the expected_* parameters aren't used when error_re_string != None,
though you have to pass them in because they come before error_re_string
in the parameter list, weird).
--
Vlad
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Mon Apr 30 23:19:17 2007