Release 0.258

Details

General Changes

  • Fix a bug in SQL functions where a query could fail with a compiler error if the function has a lambda variable with the same name as a column or function input.

  • Add Cauchy distribution CDF cauchy_cdf() and inverse CDF inverse_cauchy_cdf() functions.

  • Add SQL functions array_dupes() and array_as_dupes().

  • Add additional details to memory exceeded error messages to simplify debugging. Disabled by default. Can be enabled by setting the verbose_exceeded_memory_limit_errors_enabled session property to true.

  • Support dynamic filtering with comparison operators. Can be enabled by setting the enable-dynamic-filtering property to true. Disabled by default.

Cassandra Connector Changes

  • Improve query performance by caching metadata to avoid extra remote calls to the Cassandra server.

Elasticsearch Connector Changes

  • Fix incorrect pushdown of IS NULL predicate in Elasticsearch.

Hive Connector Changes

  • Fix a bug in Parquet dereference pushdown that caused inconsistent query results in certain cases.

  • Add support for allowing to match columns between table and partition schemas by names for HUDI tables. This is enabled when configuration property hive.parquet.use-column-names or the hive catalog session property parquet_use_column_names is set to true. By default they are mapped by index.

Iceberg Connector Changes

  • Add support for ORC files.

SPI Changes

  • Add getClientTags to ConnectorSession.

Credits

Andrii Rosa, Beinan, Chen, Grace Xin, Huameng Jiang, Jalpreet Singh Nanda (:imjalpreet), James Petty, James Sun, Jeremy Craig Sawruk, Julian Zhuoran Zhao, Junyi Huang, Marilyn Beck, Neerad Somanchi, Pranjal Shankhdhar, Roman Zeyde, Rongrong Zhong, Sergey Pershin, Shixuan Fan, Tim Meehan, Venki Korukanti, Xiang Fu, Zhan Yuan, Zhenxiao Luo, derektbrown, jiachen, lijieliang, 护城