Use pre-commit as linter & Checker ✨

[yezz123]

Based on this issue #344, I was thinking about why not use .pre-commit as a linter and checker.

Easy way to check code and fix any issue related to code style and code linting, about the rules I use only simple and most used rules for linting, SQLalchemy's team is free to change the rules or use another one for linting.

Note: This is just a simple enhancement to repository structure.

[CaselIT]

I'm personally -1 on some of these.

In particular I prefer to keep the configuration out of .pre-commit so that one is free to use the same tools with other runners.

[CaselIT]

this removes the test run, so maybe we can have this together with the current version

[CaselIT]

we use zimports so -1 here

[CaselIT]

I think it makes more sense to keep the configuration in setup.cfg

[CaselIT]

also tox is probably a better place for this

[jvanasco]

+1 to @CaselIT's suggestion on this. I generally like the idea of linting/checking on a pre-commit hook, but those checks will often need to be run on-demand during development -- and configuring them within pre-commit would require them to be duplicated in other config files.

[CaselIT]

this seems to run autoflake. zimports does mostly the same

[CaselIT]

please keep the current version. also the setting should be in pyproject.toml

[CaselIT]

not sure it's really needed. we use test for this

[CaselIT]

Thanks for the work. I've left a review.

It's probably best to wait for @zzzeek on this also so a single set of changes can be done

[bourke]

Thanks for the contribution. As @CaselIT and @jvanasco mentioned above, this PR gives pre-commit more responsibility in our tooling than we're comfortable with — especially considering that we'd prefer to keep the configurations as consistent as we can across the sqlalchemy organisation. @zzzeek suggested that the configuration for dogpile.cache would be a good reference point.

[bourke]

Hey there -- closing this for now. Let me know if you want to take it back up.