Previously, every AppKit tests used the same sharedApplication. Due to this implementation, a could not pass because a previous test made failed the current test. For instance, a test could fail because the window of the previous test resigned (just imagine a new window is the key window in the current test), and this resign could raise an error. The error was displayed for the current test thought this test was perfect !
We now instead of using sharedApplication create a new CPApplication per unit-test file in the class method setUp.
Before this fix, CPApp tried to add nil to its internal list of windows, even that nil should not be added to a CPArray. Second, that nil was actually returned by [CPApp windows] rather than an expected list of only CPWindow instances.
This fix makes [CPApp windows] return only actual windows, replaces the internal _windows[0] placeholder with CPNull, and ensures this CPNull is never actually exposed outside of the class.
Alexandre Wilhelm
Antoine Mercadal
Alexander Ljungberg
Martin Carlberg
Johannes FahrenkrugandFrancisco Ryan Tolmasky I
Klaas Pieter Annema