(BETA) Collect physical table and column statistics from a data source

(BETA) Reads the statistics the database already maintains in its own catalog, without scanning any data. Supported for StarRocks (both internal catalog for native/PIPE tables and external catalog for Iceberg tables), BigQuery, PostgreSQL, Redshift, Snowflake, Vertica, SQL Server, Azure SQL, MySQL, MariaDB and SingleStore. Statistics include row counts and data sizes, and, where the database exposes them, NDV (number of distinct values), null counts and min/max values. Returns an empty result for a data source whose database exposes no such catalog. Also returns an empty result for a data source that sets disableStatistics.