slate anvil
This is more of architectural/best practices question.
There's 2 main functions generate(...) -> Set(Action) and do(Action)
do calls generate to make sure the passed Action is valid.
So far almost all of my unit tests only concern generate (as that is 90% of my logic) and they don't test if do is processing those Actions correctly (as it is somewhat simple)
Do you think I should still add unit tests for do for those many Action cases too? Or can I just skimp out?
It'd be a lot of code. Basically every test for generate would have an accompanying test for an equivalent do if I were to test everything
sacred prism