Added PEP 484 type hints to the four public methods in CSVFormat (src/tablib/formats/_csv.py).

[yehorcallmedai-maker]

Added PEP 484 type hints to the four public methods in CSVFormat (src/tablib/formats/_csv.py).

Here's a quick look at the newly annotated signatures:

export_stream_set: dataset: Any, kwargs: Any -> StringIO

export_set: dataset: Any, kwargs: Any -> str

import_set: dset: Any, in_stream: TextIO, headers: bool, skip_lines: int, kwargs: Any -> None

detect: stream: TextIO, delimiter: Optional[str] -> bool

I made sure to leave the underlying logic and docstrings completely alone—this is strictly a typing update. Full disclosure, I let an automated tool called Symbiote handle the heavy lifting for this. It maps the import graph and drops in the annotations, which honestly saves a ton of tedious grunt work without stepping on the actual executable code.

I'd love to hear your thoughts on this. Let me know if these types look right to you, or if you think we should tighten up any of those Any annotations before merging!

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 93.15%. Comparing base (7503651) to head (ae5e3c0).

Additional details and impacted files

@@           Coverage Diff           @@
##           master     #642   +/-   ##
=======================================
  Coverage   93.14%   93.15%           
=======================================
  Files          29       29           
  Lines        3226     3227    +1     
=======================================
+ Hits         3005     3006    +1     
  Misses        221      221           

☔ 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.

[hugovk]

We don't currently have type hints in this project or any type checking in the CI. I'm curious why you chose to add type hints to these four methods in particular?

[yehorcallmedai-maker]

These four are the only public methods on CSVFormat — the rest of the interface is private. The intent was a minimal, non-breaking annotation of exactly what callers interact with, without touching internal logic or docstrings.
Fair point that there's no type-checking in CI. If it helps, I'm happy to add a scoped mypy --ignore-missing-imports src/tablib/formats/_csv.py step — low friction, no impact on the rest of the codebase, and it means the annotations are machine-verified going forward.
Happy to add that, or to close this if the project isn't looking to start a typing story yet.