pvary opened a new pull request #2054: URL: https://github.com/apache/iceberg/pull/2054
To allow existing Hive workloads to easily move to Iceberg tables we should make it easier to migrate the scripts. Currently the following types are not allowed in Hive `CREATE TABLE` statements since they do not have an exact match in Iceberg types: - TINYINT - SMALLINT - VARCHAR - CHAR We still want to keep the 1-on-1 schema mapping between Hive and Iceberg, but for the table creation we can convert unsupported types to supported ones: - TINYINT - INTEGER - SMALLINT - INTERER - VARCHAR - STRING - CHAR - STRING So this PR introduces a new configuration key: `iceberg.mr.schema.auto.conversion` which if it set to `true` then the `CREATE TABLE` query will not throw an exception, but will convert the columns instead. For example: ``` CREATE EXTERNAL TABLE not_supported_types (not_supported TINYINT); DESCRIBE not_supported_types; ---------------- not_supported | int | from deserializer ---------------- ``` ---------------------------------------------------------------- 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. For queries about this service, please contact Infrastructure at: [email protected] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
