Say some of my tests have side effects/logs/state. I don't care about the state if the tests pass, but if they fail I want to be able to see it.
Is there some way to:
* Get some identifier for a specific test run, preferably human readable, so it could be used in a log filename, database entry, or whatever other state the thing being tested might leave behind?
* Run cleanup code dependant on if the test passed?
* Output a string to the console on a failure to assist whoever is fixing the test? For example:
> mocha
screenshots are correct
test page
1) /view1
Screenshots differ from expected! See /tmp/test5512/screenshots_are_correct_test_page_view1/expected.png and actual.png. For help fixing this issue, see
http://wiki/ScreenshotMismatch
0 passing (2s)
1 failing