I have a test suite where all tests are somewhat flaky. I. e. 1-2 tests out of ~4000 fail almost every time. So I'd like to account for exactly that - rerun up to 2 tests, then fail.
How could the API look like? pytest.mark.flaky on a class/module just marks all tests as flaky - and should likely continue working this way.
I have a test suite where all tests are somewhat flaky. I. e. 1-2 tests out of ~4000 fail almost every time. So I'd like to account for exactly that - rerun up to 2 tests, then fail.
How could the API look like?
pytest.mark.flakyon a class/module just marks all tests as flaky - and should likely continue working this way.