Support for Reading Partitioned DeltaLake Data and Inferring Schema from Metadata

Why it matters

This feature enables efficient and accurate integration with DeltaLake tables by reducing the overhead of schema inference. Instead of reading data files to determine the schema, ClickHouse now reads the DeltaLake metadata, improving performance and reliability when querying partitioned DeltaLake data.

How to use it

Users can query DeltaLake tables directly in ClickHouse where the schema will be automatically inferred from the DeltaLake metadata. No additional configuration is required to enable this feature; it works transparently when accessing DeltaLake partitions.