Update pinned pytest version to 6.2.1 and fix error in test_exceptions.py::test_python_alreadyset_in_destructor #2741
Conversation
|
I think we just need to add a marker to turn this extra pytest feature off for this test. We do want this behavior. I noticed the dropping of 3.5 in the changelog. |
|
It would be nice if we could explicitly turn it off for just that test, though, and not for everything, with the blunt
|
|
Can we use the |
…k__ instead of sys.unraisablehook
|
Got something. The original built-in |
YannickJadoul
changed the title
YannickJadoul
changed the title
|
Why make it a string but still error out if unfound, PyTest... needs an if to see if that exception exists. |
tests/test_exceptions.py
Outdated
| "ignore::pytest.PytestUnraisableExceptionWarning" | ||
| ) | ||
| else: | ||
| ignore_pytest_unraisable_warning = lambda f: f # NOQA: E731 |
There was a problem hiding this comment.
Yes, I could not put a lambda, and put 2 lines def ignore_pytest_unraisable_warning(f): and return f, but then black suddenly adds an empty line after else:, so that's just as ugly.
Yeah, indeed ... :-/ Guess they're just following the |
tests/test_exceptions.py
Outdated
| @@ -50,13 +50,23 @@ def test_python_call_in_catch(): | |||
| assert d["good"] is True | |||
|
|
|||
|
|
|||
| if hasattr(pytest, "PytestUnraisableExceptionWarning"): # Python >= 3.8 and pytest >= 6 | |||
There was a problem hiding this comment.
Alternatively, we move this into the function, and directly use the warnings module. But I'm not sure whether such a filter would get reset after the test or not, this way.
|
Green! :-) Ready to be reviewed. |
| def test_python_alreadyset_in_destructor(monkeypatch, capsys): | ||
| hooked = False | ||
| triggered = [False] # mutable, so Python 2.7 closure can modify it | ||
|
|
||
| if hasattr(sys, "unraisablehook"): # Python 3.8+ | ||
| hooked = True | ||
| default_hook = sys.unraisablehook | ||
| # Don't take `sys.unraisablehook`, as that's overwritten by pytest | ||
| default_hook = sys.__unraisablehook__ |
There was a problem hiding this comment.
I take it we're happy with this?
If some case where this would be overwritten (by something else than pytest), we're ignoring it as well, ofc, but ... myeah, best I could immediately think of.
There was a problem hiding this comment.
Yes, it seems fine to me. We don't want pytest's version here, and it's a test, so nothing else should be fiddling with it.
|
Does this need to be marked 2.6.2, still? I guess it's kind of independent from the releases, but ... Guess it's mostly #2573 that needs to make sure it also installs |
|
Great! We don't really control the PyTest version if someone wants to run our tests (a requirements.txt is an internal pinning, not an external one), and a few of our tests use system PyTests, but it's good not to be (too) version specific, tests failing due to PyTest is not ideal. |
|
Do you want a changelog for this, @henryiii? |
|
Yes, I added it to the description. Here it is in setuptools changelog, for example: https://setuptools.readthedocs.io/en/latest/history.html#misc |
Great! Hadn't spotted that; thanks! :-) |
Description
As noted in #2573 (comment), pytest 6 seems to make one of our tests fail. Let's check this, and then figure out if/how to fix that.
Suggested changelog entry: