It would be great if we can implement the failed rows processor for Soda Spark as well: https://github.com/sodadata/soda-sql/blob/main/core/sodasql/scan/failed_rows_processor.py