hankfanchiu opened a new pull request #2484: URL: https://github.com/apache/hive/pull/2484
### What changes were proposed in this pull request? On the Hive MetaStore server's side of the `HiveMetaStoreClient#listTableNamesByFilter()` API, conditionally manipulate the string filter by replacing any `=` comparison on a table parameter with a `LIKE` operator. This replacement is only performed if: 1. The database product backing the Hive MetaStore is Derby, 2. The `=` comparison is on a table parameter, i.e. the `"TABLE_PARAMS"."PARAM_VALUE"` column. ### Why are the changes needed? HIVE-12274 changed the type of the `"TABLE_PARAMS"."PARAM_VALUE"` column to `CLOB` for Derby and Oracle, e.g.: https://github.com/apache/hive/blob/7d4134e7fe9bfb8b8aa3344a9ae72f4f36c98b2c/metastore/scripts/upgrade/derby/039-HIVE-12274.derby.sql#L8-L12 With the type change, invoking `ObjectStore#listTableNamesByFilter()` given a `=` comparison -- on a table parameter stored in Derby -- fails with the following exception: ``` ERROR 42818: Comparisons between 'CLOB (UCS_BASIC)' and 'CLOB (UCS_BASIC)' are not supported. Types must be comparable. String types must also have matching collation. If collation does not match, a possible solution is to cast operands to force them to the default collation (e.g. SELECT tablename FROM sys.systables WHERE CAST(tablename AS VARCHAR(128)) = 'T1') ``` Without reverting the column type change, we add support for `=` comparison on table parameters in Derby. ### Does this PR introduce _any_ user-facing change? Yes, invoking `HiveMetaStoreClient#listTableNamesByFilter()` with a `$param_key = '$param_value'` filter now succeeds for Derby databases. ### How was this patch tested? -- 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]
