February 2026: Focused on correctness and reliability of catalog management in Spark SQL. Delivered a targeted bug fix for SET CATALOG to properly handle catalog names containing special characters, ensuring correct resolution during analysis. Implemented quoting with backticks and escaping, added focused tests, and preserved backward compatibility while expanding support for complex catalog names. The change reduces runtime failures in multi-catalog environments and improves user experience when working with special-character catalog names.

January 2026 monthly summary for the apache/spark repository focused on extending the SET CATALOG syntax. Delivered support for foldable expressions in SET CATALOG, enabling dynamic catalog name resolution directly in SQL without wrapping in IDENTIFIER(). This enhances usability, reduces boilerplate, and allows expressions such as CONCAT('test', 'cat') and CAST('my_catalog' AS STRING) to be evaluated at runtime. The change preserves case-sensitivity for all input forms and explicitly does not support multi-part identifiers. The patch was validated with unit tests in DataSourceV2SQLSuite, covering foldable expressions, case sensitivity, session-variable interactions, and negative scenarios (non-foldable, null, and non-string types). tied to SPARK-55155 and merged commit 3de3d81d01f66b9547a55000206b740bafcd8878.