Hi all, I would like to collect coverage statistic for my twisted code with trial and was wondering how to achieve this. I run something like: $ trial --coverage testdir/*_test.py all tests are green and i can see coverage files in _trial_temp/coverage/. After looking around a bit i learn that the python trace module is used under the hood and my assumption then is that i must use similar commands as those that are shown here: http://docs.python.org/library/trace.html I had to look at the source code to find this out. It would be nice if it was mentioned in the documentation somewhere. So i try to construct a command like this: python -m trace --summary --report --coverdir _trial_temp/coverage/ But it needs --file which seems not to be enabled by trial. I then make some changes in the trial script: --- trial.py.orig 2012-03-30 06:32:08.627758872 +0200 +++ trial.py 2012-04-14 04:50:05.416382063 +0200 @@ -162,7 +162,8 @@ directory specified by the I{trial-temp} option. """ import trace - self.tracer = trace.Trace(count=1, trace=0) + countfile="countfile.txt" + self.tracer = trace.Trace(count=1, trace=0, infile=countfile, outfile=countfile) sys.settrace(self.tracer.globaltrace) and i can finally see a summary if i remember to add the --missing option: python -m trace --report --summary --missing --file countfile.txt -- coverdir=_trial_temp/coverage/ I assume i am missing something obvious, because it seems strange that one would have to change the trial script to achieve this. How do other people create such summaries? --- Øystein