TST: find solution for extension/tests_numpy.py (PandasArray base extension tests)

[jorisvandenbossche]

We currently run the base extension tests also for the PandasArray (to have coverage of this array class), but because we actually don't allow to store this in a Series/DataFrame, the tests do some patching to workaround this:

pandas/pandas/tests/extension/test_numpy.py

Lines 36 to 54 in dec7d21

	def allow_in_pandas(monkeypatch):
	"""
	A monkeypatch to tells pandas to let us in.
	By default, passing a PandasArray to an index / series / frame
	constructor will unbox that PandasArray to an ndarray, and treat
	it as a non-EA column. We don't want people using EAs without
	reason.
	The mechanism for this is a check against ABCPandasArray
	in each constructor.
	But, for testing, we need to allow them in pandas. So we patch
	the _typ of PandasArray, so that we evade the ABCPandasArray
	check.
	"""
	with monkeypatch.context() as m:
	m.setattr(PandasArray, "_typ", "extension")
	yield

However, this regularly gives problems, see eg #39991 (comment) or #39722 (comment), where for some reason those PandasArrays do / don't get preserved in a DataFrame/Series due to some other code changes that impact this, and the tests are not written to deal with it.
(one example: extract_array will convert a pandas Series with numpy dtype into a PandasArray, when the patch is enabled. This can then cause the result to include the PandasArray EA, while the expected result is not created that way)

So we should try to find a solution for this (as it gets quite annoying to deal with it just to get those tests passing). But I don't directly know what the best solution would be.
Some ideas:

• Simply remove tests/extension/test_numpy.py and thus don't run the base extension interface tests for PandasArray. We still have tests/arrays/test_numpy.py for some other tests. The easiest solution, but would also remove test coverage.
• In some way only run the tests in tests/extension/test_numpy.py that only deal with the array, and not with the array stored in DataFrame/Series.

cc @TomAugspurger @jbrockmendel

[jbrockmendel]

In some way only run the tests in tests/extension/test_numpy.py that only deal with the array, and not with the array stored in DataFrame/Series.

This seems reasonable to me. Unless a) we have some plan to start allowing PandasArray in Blocks or b) we think some significant portion of the PandasArray APi is only tested in the Series/DataFrame tests (in which case we should likely refactor the tests to include more direct testing)