mutt --help will give usage information. In short, running
mutt with no arguments or
mutt testall will run both the python
tests, and running
mutt testpy will run only the python tests.
These arguments correspond to the following manifests:
- testall : all-tests.ini
- testpy : pythontests.ini
- testjs : jstests.ini
About the tests
In general, tests should be good usage examples. Try to write a test that illustrates an issue! While sometimes this is not possible, it is a good aspiration to live by.
The manifests use the manifestparser format.
Run the tests before pushing to master or any branch of mozilla . What is lost in time here is more than made up for vs. hours of regression hunting.
The mutt tests should pass all the time on all platforms of interest! If this is not the case, there is a problem. If a test fails, does it fail without your changes?
In either case, the following actions are advised:
if you can, fix the test case. Please. jhammel will personally buy you a beverage of choice
#mozmillon irc://irc.mozilla.org/ and mention what you observe
if the failure is not caused by your changes, file a bug detailing what you're observing
if the test cannot be fixed immediately but should not be deprecated (so, the last case resort), remove the test from the appropriate test manifest. Note this in the bug. It is a higher value (at least for now) to have 100% passing tests than to have higher test coverage.