[fhgffy]

Describe your changes
When no CSV columns are supplied, to_csv() converts the first row's keys to strings before looking up row values. Integer keys therefore produce empty cells: {1: "first", 2: "second"} is exported as 1,2\n,\n.

Keep the original keys for lookup and sort them by their string representation. Inferred headers keep their string labels, including None, while explicit columns keep their existing writer behavior. Regression cases cover integer keys, mixed string/integer keys, None keys, explicit None columns, and output without a header row.

Related issue
Found while checking serializer round trips; no existing issue found for this case.

Checklist before requesting a review

• I have performed a self-review of my code.
• I have added tests for the proposed changes.
• I have run the tests and there are not errors.

Validation on Python 3.14: four regression cases fail on the original source; related tests pass (18 tests); the full suite passes (928 tests, 1 skipped); all pre-commit hooks including Ruff/mypy pass; wheel and sdist builds pass.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 98.51%. Comparing base (8c3371f) to head (686eb62).

Additional details and impacted files

@@           Coverage Diff           @@
##             main     #622   +/-   ##
=======================================
  Coverage   98.51%   98.51%           
=======================================
  Files          64       64           
  Lines        2430     2431    +1     
=======================================
+ Hits         2394     2395    +1     
  Misses         36       36           

Flag	Coverage Δ
unittests	98.51% <100.00%> (+<0.01%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ 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.