Worked on the apache/parquet-java repository to deliver deterministic sorting of encoding formats within the Parquet metadata converter, addressing nondeterminism in how encodings are retrieved and consumed downstream. Implemented the sorting logic in Java within the ParquetMetadataConverter, accompanied by a dedicated unit test to verify consistent ordering. Refactored existing test cases to resolve formatting violations, thereby improving CI reliability and overall code quality. This contribution enhanced metadata reproducibility and stability for downstream processing, supporting more predictable integration and developer workflows. Collaborated closely with co-authors throughout the review process, demonstrating a focus on software development best practices and robust unit testing.