On Tue, Mar 3, 2020 at 2:18 PM Sven Baars sbaars@codeweavers.com wrote:
On 03-03-2020 14:06, Matteo Bruni wrote:
Hi Sven,
I'm attaching a diff to be applied on top of this patch. Does it look okay to you? In particular I'm interested if you have more tests not yet upstream.
Hi Matteo,
I have some more tests, but only for DrawText. I send those with the patches that fix the tests, because otherwise they would add dead code.
If you apply a similar diff to this one to the first DrawText patch (2/5), however, I expect the tests that are already present to fail (I believe those tests also represent actual applications). For consistency I did the same thing for PreloadText. I could add some tests to check if it also fails for PreloadText if you like.
Cool, that's what I was looking for, thanks. Yes, similar changes to DrawTextA() cause a failure for the "", -1 test (in particular because of removing the conditional - 1 to countW). I think I still like it more like this though (after fixing the test failure, of course). Anyway, I'll review the other patches later.
BTW I don't think you can test that for PreloadTextA() since the function doesn't return a character count or anything observable.