[WIP] Mypy Check Bug Fix

[daralynnrhode]

A fix for the mypy bug and any needed changes

Overview

Closes #710

Ensures that mypy will only check files within imap_processing/ etc. Changes to make the pre-commit mypy check and regular mypy check align.

Currently, I am being pushed in the direction to think it is a recursion difference that is causes the difference in errors for the two checks. I am trying to track down where this is occurring.

For Reference:

mypy . ran from imap_processing/ produces

pre-commit run --all ran from imap_processing/ produces

To Note; all errors found by running the pre-check are present when the mypy check is ran by itself

New Dependencies

New Files

Deleted Files

Updated Files

• pyproject.toml
• .pre-commit-config.yaml
• codice_l1a.py

Testing

[greglucas]

Looks like you need to merge or rebase on upstream/dev to get the latest changes before the tests can run.

[daralynnrhode]

With the most recent commit I am still having errors produced when just mypy is ran.

[daralynnrhode]

Now back to the point of pre-commit passing but the local mypy check produces errors. I am hedging toward this being caused because of a difference in the recursive calls for mypy when it is called individually versus in pre-commit.

In the mypy documentation it is stated that submodules found in what is listed in packages will be checked, so the pyproject is covering all files within imap_processing and tools. Shouldn't the pre-commit-when it runs the mypy check-check the pyproject for how mypy should be configured?

[daralynnrhode]

Turns out since the pre-commit runs its check in an isolated virtualenv some things may be lost.
https://github.com/pre-commit/mirrors-mypy

I believe the dependencies need to be listed for mypy but am at a loss as to what all those dependencies should be.

[greglucas]

I think that for all of the numpy typing we need to import their typing module, and then use npt.NDArray. Saying that we are returning np.ndarray seems incorrect, so I think we should change those, rather than add these explicit type annotations throughout the code.

import numpy.typing as npt

def func() -> npt.NDArray

https://stackoverflow.com/questions/66349242/specific-type-annotation-for-numpy-ndarray-using-mypy