huan233usc commented on issue #2050: URL: https://github.com/apache/iceberg-rust/issues/2050#issuecomment-4595802411
Question here: Do we want to support `CREATE TABLE PARTITIONED BY` or `CREATE EXTERNAL TABLE PARTITIONED BY`？ These two readings of the title imply very different work: - **`CREATE TABLE foo (...) PARTITIONED BY (...)`** — what Spark/Trino/Flink users expect. But today DataFusion's planner rejects `partition_by` on plain `CREATE TABLE`, and `CreateMemoryTable` + `SchemaProvider::register_table` have no slot for partition info. So this requires upstream apache/datafusion changes first. - **`CREATE EXTERNAL TABLE foo STORED AS ICEBERG LOCATION '...' PARTITIONED BY (...)`** — reachable today via `IcebergTableProviderFactory` (the path PR #2022 took). But this only loads existing tables, can't create new ones, and `CREATE EXTERNAL TABLE` is essentially a Hive-legacy construct that no other modern Iceberg integration uses. cc @CTTY @liurenjie1024 @charlesdong1991 for thoughts -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
