Martin Pitt

    def tearDown(self):
        [...]
        # collect log, run() shows it on failures
        with open(self.daemon_log.name) as f:
            self.log_output = f.read()

    def run(self, result=None):
        '''Show log output on failed tests'''

        if result:
            orig_err_fail = result.errors + result.failures
        super().run(result)
        if result and result.errors + result.failures > orig_err_fail:
            print('\n----- daemon log -----\n%s\n------\n' % self.log_output)

11 Comments

• #1 by Jonathan Lange on 2012/10/11 - 12:19

  Zitieren

  

  Oh man, we’ve got this _so_ covered, but you need to use some other dependencies.

  testtools has an API called addDetails, which lets you attach arbitrary content to a test’s result. These results are then shown in the output.

  http://testtools.readthedocs.org/en/latest/for-test-authors.html#details

  That solves the core problem.

  However, if you want to get even fancier, you can use fixtures: . They also support the same “details” API. What you’d do is make a Fixture subclass for your daemon and have that class add the log. Then, in your test::

  self.useFixture(GnomeSettingsDaemon())

  And it will take care of showing the log and shutting down the daemon.

  This has the added advantage of making the daemon set up and tear down re-usable without subclassing.

• #2 by Marcel Stimberg on 2012/10/12 - 01:14

  Zitieren

  

  If you’d be using the nose testing framework (and why would you use anything else ?), you’d have this behaviour by default:
  https://nose.readthedocs.org/en/latest/plugins/logcapture.htm

• #3 by Marcel Stimberg on 2012/10/12 - 01:16

  Zitieren

  

  Ups, that link was missing an l in the end…:
  https://nose.readthedocs.org/en/latest/plugins/logcapture.html

• #4 by pitti on 2012/10/12 - 06:12

  Zitieren

  

  Marcel,

  indeed I do use nose for some of my Python projects; these are tests for gnome-settings-daemon which uses autotools (and is otherwise a C project). But doesn’t this only capture things you already print to stdout/err? What I want is to only show the daemon log for a test that fails.

• #5 by Ulrich Eckhardt on 2012/10/15 - 07:04

  Zitieren

  

  You can print() things unconditionally and tell unittest.runt() to buffer that output. In case of success, the output is discarded, otherwise it is displayed along with the fault.

  Uli

• #6 by pitti on 2012/10/15 - 07:08

  Zitieren

  

  Ah, do you have an example how to do that?

• #7 by Ulrich Eckhardt on 2012/10/16 - 21:47

  Zitieren

  

  Sorry, should have been main(), not runt(), which was a typo anyway.

  From http://docs.python.org/library/unittest.html, it’s just “unittest.main(buffer=True)” and you won’t see any of the output of setUp, test and tearDown unless the thing fails. Alternatively, you can pass -b or –buffer on the commandline.

  Should be enabled by default, if you ask me.

  Uli

• #8 by pitti on 2012/10/17 - 08:38

  Zitieren

  

  Ah, thanks Ulrich!

• #9 by Marcel on 2012/10/17 - 23:28

  Zitieren

  

  Sorry for not following up on my comment. The logcapture extension does capture log messages emitted using Python’s logging system. I assumed that was how you generated the log file but I only now realized that it is about a different log file. You could still always “log the log” (i.e. do something like logger.debug(log_output) but then Ulrich’s solution is probably easier.

• #10 by edin on 2013/04/11 - 15:36

  Zitieren

  

  Hi,
  Do yuo know how to repeat only failed tests? Some call in tearDown…

• #11 by pitti on 2013/04/12 - 04:19

  Zitieren

  

  You can run the test suite with explicit test names (tests/run MyClass.test_foo) if you want to debug a particular test and only run that one.