Surface.get_locked test

[PeanutbutterWarrior]

Test for #1803

[MyreMylar]

Looks pretty solid, I suppose we could also try

• blitting a locked surface onto another one.
• blitting onto a locked surface and confirming it produces an error.

(you can check for assertions being raised with self.assertRaises and passing the function name and it's parameters as separate arguments, see examples here:

https://github.com/pygame/pygame/blob/master/test/blit_test.py#L135)

[PeanutbutterWarrior]

Sorry, but how does that test .get_locked()? As another check that a surface is locked?

[MyreMylar]

Sorry, but how does that test .get_locked()? As another check that a surface is locked?

Yeah, my thinking was just a double check that the flag was accurately reporting what was happening deeper in the code somewhere. I mean probably the code right now is just checking for the flag in blit before raising an assert so we would be just in a circle, but perhaps at some point in the codes life it might check for locked-ness of a surface in a different way internally in SDL or pygame and the flag we check with get_locked() could get out of sync with what is going on.

The blit code is a big old mess of code so I could definitely see it :)

Honestly, I think this kind of testing could easily fit in the unit test for lock() (when that exists) - checking that surfaces locked there both can't be blitted on and report that they are locked with get_locked() so we could just punt it over to there if you think it makes this test clearer?

[PeanutbutterWarrior]

Yes, I think this would work as a test for lock(), then for get_locked() it will check against the blit check, rather than the expected value.

[MyreMylar]

Then it all looks good to me, I'll make a mental note for the .lock() unit test. 👍 🍾

[MyreMylar]

New changes also look good! :)

[illume]

👍 thanks