I'm running Travis and Coveralls to run integration tests. It appears Coveralls is giving me a much higher rating than it should; I have only written tests for a handful of source files in my repository, and those are the only files that get checked for coverage (hence I get 95% coverage..). It appears as if files that do not have any tests specified do not get checked.
How does Coveralls select which files to check?
Perhaps my repository is a bit unusual, in that it contains several Python scripts that each have their own entry point. I am currently only testing one entry point, leaving the rest of the code completely untouched (and not even imported).
My travis.yml file simply executes nose2, which runs tests for this one entry point.
