Writing some tests for a plugin using TestKit. For some reason, TestKit thinks my tasks are up-to-date when those tasks’ output file does not exist.
It works perfectly fine in a “real” project using the plugin, but “fails” in the TestKit tests because I explicitly check for TaskOutcome.SUCCESS
while TestKit says the tasks are TaskOutcome.UP_TO_DATE
.
I very much dislike publishing code if I cannot properly test it. Yes the external project could act as the verification, but I would first have to publish the plugin (to the portal or a Maven repo) first which defeats the purpose of having unit/functional tests.
So am I missing something? Is this just a quirk in TestKit? How do I resolve that?
FWIW as I understand it, it should not be a daemon issue as I have withDebug(true)
set