[
https://issues.apache.org/jira/browse/HIVE-679?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12742153#action_12742153
]
Bill Graham commented on HIVE-679:
----------------------------------
Thanks for the comments and for catching those tabs.
We could use CHANGES.txt for the driver version, but I think the solution
should be done at build time, instead of runtime. It seems like an undesirable
dependency to have a txt file bundled in the hive_jdbc.jar that has to be read
at runtime somehow. If it was done at build time though, we'd need to do some
token substitution in the java source before compiling to insert the version
taken from CHANGES.txt. Also not the prettiest, but I think the end result
would be better. What do you think?
I'm importing serdeConstants for the getColumnTypeName method, which converts
Thrift types to Hive types. I debated hardcoding the Hive types in
HiveResultSetMetaData to avoid the dependency, but since serde is already
required by hive_jdbc elsewhere, I used serde.Contants.
> Integrate JDBC driver with SQuirrelSQL for querying
> ---------------------------------------------------
>
> Key: HIVE-679
> URL: https://issues.apache.org/jira/browse/HIVE-679
> Project: Hadoop Hive
> Issue Type: New Feature
> Components: Clients
> Reporter: Bill Graham
> Assignee: Bill Graham
> Attachments: HIVE-679.1.patch
>
>
> Implement the JDBC methods required to support querying and other basic
> commands using the SQuirrelSQL tool.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.