Add unit tests

[kaitj]

Another large PR - this time adding unit tests to niftyone. Tried to summarize everything below, but may have missed a few things:

• Updated the use of the importlib.resources - the usage in _annotate.py was deprecrated in py39.
• Fixed a typing incompabilty issue in _pad.py
• Excluded dunders from the testing and coverage, as well as anything the the typing.py files.
• Reorganized the tests to (1) reflect unit tests (if there are other end-to-end tests in the future) and (2) reflect the source code organization.
• Added fixtures, some of which are global
• Mocked some of the objects / calls where the testing went a little more in depth
• For figures, only tested to make sure the figures were generated (and saved)
• Where possible, I've added tests for both 3D and 4D images (and just a numpy array if appropriate)

I think some of these tests could probably go even more in depth than they are, but this brings the coverage up and should hopefully help to catch issues as any code gets refactored and extended. There was 1 line (I think in render?) where it called mux_one from the av library, that I couldn't figure out how to test / cover.

[codecov[bot]]

Codecov Report

All modified and coverable lines are covered by tests :white_check_mark:

Project coverage is 99.84%. Comparing base (72410f5) to head (adbcd66).

:exclamation: Current head adbcd66 differs from pull request most recent head 2f7bff3. Consider uploading reports for the commit 2f7bff3 to get more accurate results

Additional details and impacted files

```diff @@ Coverage Diff @@ ## main #8 +/- ## ========================================== + Coverage 5.87% 99.84% +93.97% ========================================== Files 24 16 -8 Lines 1022 631 -391 ========================================== + Hits 60 630 +570 + Misses 962 1 -961 ```

:umbrella: View full report in Codecov by Sentry.
:loudspeaker: Have feedback on the report? Share it here.