On Tue Jan 2 18:37:46 2024 +0000, Zsolt Vadász wrote:
the first patch is supposed to pass the tests without the second applied
so you need a todo_wine clause in from of the failing tests in first
patch, and to remove it in second patch when the fix is applied i'm not sure i understand. the failing tests are unrelated to my patch. could you rephrase this part?
First patch is fixing the implementation, second patch adds tests so there's no need to add todo_wine.
I would have changed existing tests to add case-sensitive case but adding a dedicated function is also OK.
Environment variables are not case-sensitive on Windows. I think it's OK to assume that C-runtime is not doing it differently. I think we can also add a test for putenv mainly to ensure the implementation is correct in this regard.
The patch looks good for me.