Fix broken CI: add pre-commit checks for better CI

[jeking3]

It looks like pre-commit was added as a requirement for CI.

This caught a few issues, like having duplicate named test methods, so I fixed those up.

Description

I added a reasonable pre-commit configuration with black, flake8, and reformatting.

Related Issue

This fixes #869

Motivation and Context

How Has This Been Tested?

Screenshots (if appropriate):

Types of changes

• Bug fix (non-breaking change which fixes an issue)
• New feature (non-breaking change which adds functionality)
• Breaking change (fix or feature that would cause existing functionality to change)

Checklist:

• I have run the make format command to format my code (well, pre-commit...)
• My change requires a change to the documentation.
• I have updated the documentation accordingly.
• I have read (and updated) the CONTRIBUTING document.
• I have added tests to cover my changes.
• I have added my name and/or github handle to AUTHORS.rst
• I have added my change to CHANGES.rst
• All new and existing tests passed.

[codecov]

Codecov Report

Merging #896 (0c2457f) into master (83ee3db) will not change coverage.
The diff coverage is 100.00%.

@@           Coverage Diff           @@
##           master     #896   +/-   ##
=======================================
  Coverage   97.83%   97.83%           
=======================================
  Files          22       22           
  Lines        1062     1062           
  Branches      162      162           
=======================================
  Hits         1039     1039           
  Misses         10       10           
  Partials       13       13           

Impacted Files	Coverage Δ
simple_history/__init__.py	100.00% <ø> (ø)
simple_history/manager.py	97.67% <ø> (ø)
...delwithcustomattronetoonefield_options_and_more.py	100.00% <ø> (ø)
simple_history/registry_tests/tests.py	96.03% <100.00%> (ø)

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 83ee3db...0c2457f. Read the comment docs.

[jeking3]

@jazzband/django-simple-history Team, CI has been broken for days. This PR fixes it, please, let's do something to make this project more efficient? All PRs require a review and virtually nobody is doing reviews.

[tim-schilling]

I'm fine with reordering imports. It's the single import per line that's too verbose for my personal tastes. Though I suppose that's why we use a linter.

[jeking3]

Reasoning around splitting comma-separated imports: https://github.com/asottile/reorder_python_imports#why-this-style
I did not disable this as of yet, but I did request a re-review so you would see this.

[tim-schilling]

Thanks for sharing that @jeking3, I now understand the reasoning. I'm still on the fence regarding it, but I'm a curmudgeon.

[jeking3]

I'll remove that from pre-commit and undo the changes, and the churn should be much less on this PR. What I really want is to get CI running again so we can clear up 4-5 PRs, and hopefully get the as_of PR code reviewed.

[macro1]

It looks like adding a blank pre-commit configuration will also get CI running again (such as the one I have in #900). I don't see what added that check to begin with..

[jeking3]

I removed the utf-8 stamping and import reordering that was previously there.

This exercise did catch two tests identically named, so I fixed that up. A few cases where assert was used instead of self.assertSomething(). Bare asserts are generally to be avoided because they are not guaranteed to run depending on the runtime environment. That was caught by bandit.

@tim-schilling take a look if you have a chance, I think you will find this more palatable.

[jeking3]

I realized I had left a couple changes in runtests.py from the import reordering and utf-8 stamping before yesterday, so I just fixed that up.

[jeking3]

@tim-schilling take a look if you have a chance, I think you will find this more palatable.

[tim-schilling]

Looks good!