Why doesn't "coverage.py run -a" always increase my code coverage percent? -
i have gui application trying determine being used , isn't. have number of test suites have run manually test user interface portions. run same file couple of times "coverage.py run file_name -a" , different actions each time check different interface tools. expect each time ran -a argument, increase code covered line count coverage.py (at least unless new files pulled in). however, gives lower code coverage after additional run - causing this?
i not editing source between runs , no new files being pulled in far can tell. using coverage.py version 3.5.1.
that sounds odd indeed. if can provide source code , list of steps reproduce problem, i'd take @ it: can create ticket here: https://bitbucket.org/ned/coveragepy/issues
Comments
Post a Comment