[
https://issues.apache.org/jira/browse/HIVE-559?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Min Zhou updated HIVE-559:
--------------------------
Issue Type: Sub-task (was: New Feature)
Parent: HIVE-576
> Support JDBC ResultSetMetadata
> ------------------------------
>
> Key: HIVE-559
> URL: https://issues.apache.org/jira/browse/HIVE-559
> Project: Hadoop Hive
> Issue Type: Sub-task
> Components: Clients
> Reporter: Bill Graham
> Assignee: Min Zhou
>
> Support ResultSetMetadata for JDBC ResultSets. The getColumn* methods would
> be particularly useful I'd expect:
> http://java.sun.com/javase/6/docs/api/java/sql/ResultSetMetaData.html
> The challenge as I see it though, is that the JDBC client only has access to
> the raw query string and the result data when running in standalone mode.
> Therefore, it will need to get the column metadata one of two way:
> 1. By parsing the query to determine the tables/columns involved and then
> making a request to the metastore to get the metadata for the columns. This
> certainly feels like duplicate work, since the query of course gets properly
> parsed on the server.
> 2. By returning the column metadata from the server. My thrift knowledge is
> limited, but I suspect adding this to the response would present other
> challenges.
> Any thoughts or suggestions? Option #1 feels clunkier, yet safer.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
