Add is_nan aliases

[BharatDeva]

Closes #1235

Rationale for this change

The Python API currently exposes is_null() with snake_case naming, but the NaN check is only available as isnan(). Adding is_nan() gives users a casing-consistent spelling while preserving the existing isnan() API.

What changes are included in this PR?

This adds datafusion.functions.is_nan(...) as an alias for isnan(...) and adds Expr.is_nan() as the corresponding expression method. Existing isnan() behavior is unchanged.

The PR also adds coverage for both the function-level alias and the expression method.

Are there any user-facing changes?

Yes. Users can now call functions.is_nan(col(...)) and col(...).is_nan() in addition to the existing isnan spelling. This is additive and does not remove or deprecate the current API.

Testing

uvx pre-commit run --files python/datafusion/functions/__init__.py python/datafusion/expr.py python/tests/test_functions.py python/tests/test_expr.py

Result: passed.

I also ran a WSL canary against the published datafusion==54.0.0 wheel with the modified wrapper files overlaid:

is_nan aliases match isnan output: [False, True, None]

Note: a full Windows uv run pytest is blocked on this machine because building nanoarrow==0.8.0 requires a local C/C++ compiler.

[kosiew]

@BharatDeva
Thanks for working on this!
I took a look and did not find any blocking issues. I have one small suggestion that could make the generated documentation a little clearer.

[kosiew]

Nice addition! Since is_nan is a direct alias of isnan, it might be helpful to make that relationship a little more obvious in the docstring. A short See Also reference to isnan, or reusing the existing isnan example, would make the generated help a bit easier to navigate.