fix: move literals to types.py

[jrycw]

Hello team,

This PR provides an initial implementation for #526.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 91.18%. Comparing base (ae0204c) to head (0b79b30).
⚠️ Report is 44 commits behind head on main.

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #672      +/-   ##
==========================================
+ Coverage   91.17%   91.18%   +0.01%     
==========================================
  Files          47       48       +1     
  Lines        5462     5471       +9     
==========================================
+ Hits         4980     4989       +9     
  Misses        482      482              

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[rich-iannone]

LGTM!

[jrycw]

Could the team help review which literals (if not all) should be included in the __all__ in __init__.py?

[jrycw]

Initially, I considered exposing all the literals directly in __init__, but I realized it might be better to expose the types instead, similar to what we did with the vals module.

[machow]

Thanks for this PR -- we're actively discussing types right now. No need to update this PR right now. I'll be sure to come back with more fleshed out thoughts, if that's okay!

Background

We'll be working on a extension package for Great Tables (with extra nanoplots, themes, etc.. that we don't want to load up the main package too hard with), so this piece will be super relevant.

Two big pieces for us to think about:

• Organizing types: how should type annotations be organized and communicated in types.py?
• Exposing types: do we want a public types module right now?

I feel like the polars._types submodule has been a helpful reference. It looks like they deprecated their polars.type_aliases submodule for now, so it might be good to marinate this feature for a bit.

[machow]

We're sprinting to merge and clear many of the outstanding PRs. I'm going to close for now, but I marked the linked issue as an epic, and definitely think it'll be good for us to come back to. Thanks for taking the time to work on this, this is mostly an issue of us getting things out the door!