Resolve typing complexity in tests and inventory

[davidverweij]

Due to the unknown payloads from the inventory API, typing becomes more complex. As such, the response can return Any type:

async def response(path: str, params: str = None) -> Any

This is further complicated in tests with the use of a monkeypatch, e.g. in tests > consumer_tests > routers > conftest.py:

@pytest.fixture(scope="function")
    def response_row():
    ...

becomes something like

@pytest.fixture(scope="function")
    def response_row() -> Generator[Dict[str, Any], None, None]:
    ...

This however fails typing as unkown methods are called on the Generator (e.g. client.get()) - which is used with the monkeypatch. Not to mention the 'private' methods in those tests and the required typing for their returned values.

At this moment, I've resorted to exclude these tests from typing checks by mypy, i.e. in mypy.ini:

[mypy-consumer_tests.routers.conftest]
ignore_errors = True

[mypy-consumer_tests.routers.test_inventory]
ignore_errors = True

I've also excluded flake8 linting for error code AAA05 as it conflicts with black introducing a white line between private methods in an Act block.

Next steps

Ideally this typing is corrected to prevent any coding errors, but priority is low.