Worked on enhancing arXiv handling in the JabRef/jabref repository by implementing a feature that separates the arXiv prefix from the EPRINT field, enabling more precise data normalization for bibliographic entries. Developed robust parsing logic in Java to accommodate a wide range of input scenarios, including corrupt or malformed data, and created targeted unit tests to ensure reliability and prevent regressions. Focused on data validation and comprehensive test coverage, the work improved the quality and consistency of arXiv metadata, facilitating cleaner downstream indexing and search while reducing manual data cleanup for researchers who rely on accurate bibliographic information.