Worked on the apache/spark repository to enhance reliability and user experience for recursive file lookups in partitioned data sources. Addressed a longstanding issue by replacing a generic IllegalArgumentException with a dedicated, user-facing AnalysisException, introducing a new error class to standardize error messaging and improve downstream handling. Leveraged Scala for backend development, focusing on robust error handling and comprehensive unit testing. Updated error reporting infrastructure and expanded test coverage to ensure the new error condition is surfaced clearly at runtime. This work laid the foundation for improved monitoring and support visibility in partitioned data scenarios, emphasizing maintainability and user clarity.