fix: Explicitly specify empty_as_null for explode

[Oliver Borchert (borchero)]

Motivation

This fixes nightly CI: https://github.com/Quantco/dataframely/actions/runs/28209348571/job/83566986568.

[Copilot]

Pull request overview

This PR fixes a nightly-CI breakage in the experimental schema inference code by making Polars Series.explode behavior explicit during nested-type (List/Array) dtype argument inference.

Changes:

• Explicitly pass empty_as_null=False to series.explode(...) when inferring inner dtypes for pl.List.
• Explicitly pass empty_as_null=False to series.explode(...) when inferring inner dtypes for pl.Array (keeping shape=... handling intact).

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 100.00%. Comparing base (50e18a8) to head (340a1c5).
⚠️ Report is 1 commits behind head on main.

Additional details and impacted files

@@            Coverage Diff            @@
##              main      #356   +/-   ##
=========================================
  Coverage   100.00%   100.00%           
=========================================
  Files           56        56           
  Lines         3458      3463    +5     
=========================================
+ Hits          3458      3463    +5     

☔ View full report in Codecov by Harness.
📢 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.

[Oliver Borchert (borchero)]

Ah polars-minimal fails. Andreas Albert (@AndreasAlbertQC) should we bump or how should we handle this?

[Andreas Albert (AndreasAlbertQC)]

Would be a shame to give up compat with older versions over this. I feel like we just have to bite the bullet and pass / not pass the argument based on polars.__version__