Add explicit support for `unittest.mock.Mock` in monkeypatch fixture about pytest HOT 3 OPEN

we should mirror the mock apis with a mock prefix and forego selective undo, people can use the monkeypatch.context context manager

from pytest.

An alternative solve, would be to update pytest-mock to use monkeypatch, so that it has similar patching semantics as monkeypatch instead of using unittest patch semantics.

Could you elaborate on this? How does pytest-mock not already do what you want? I'm -1 on introducing a different way to do the same thing into the core when we already have a well maintained and popular plugin that solves the same problem.

from pytest.

Actually thinking about it more carefully - I was wrong to suggest that pytest-mock could be updated to solve this. Extending pytest-mock to allow patching with non-mock objects wouldn't make a lot of sense.

The issue with using pytest-mock as the only means of patching is that it defaults to using MagicMock. MagicMock can be problematic / difficult to work with (it has a lot of 'magic' that leads to some pitfalls for developers). As a result I think a lot of teams end up in a situation where they use a combination of monkeypatch (for simpler cases) as well as one of either pytest-mock or unittest.mock (when they need more fully featured 'mocking').

It forces developers to understand the semantics of two libraries in order to patch objects in unit tests. We shouldn't need more than one tool to solve one problem (patching).

IMO developers should have a single way of patching (monkeypatch) with better support for patching with MagicMock/AsyncMock for the cases where that is needed.

If the answer was "just use pytest-mock" or any other plugin - I would be happy. But I think we currently exist in a limbo where you end up using multiple tools to patch objects when writing tests in Pytest.

from pytest.