[
https://issues.apache.org/jira/browse/PIG-590?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12673117#action_12673117
]
Olga Natkovich commented on PIG-590:
------------------------------------
I committed the patch. Everything looked really good. I found one thing that
would be nice to fix. I had bad UDF and got the error below:
ERROR 2078: Caught error from UDF:
org.apache.pig.test.udf.evalfunc.BadUdf[Exception occurred in
BadUdf::exec.org.apache.pig.backend.executionengine.ExecException: ERROR 1072:
Out of bounds access: Request for field number 2 exceeds tuple size of 2]
I think it is a bit confusing to have 2 error codes in the same error message.
My proposal would be, for UDFs that play by the rules and include error codes,
to use that error code in the output; for others, use our own generic error
code.
> Error handling phase three
> --------------------------
>
> Key: PIG-590
> URL: https://issues.apache.org/jira/browse/PIG-590
> Project: Pig
> Issue Type: Sub-task
> Components: impl
> Affects Versions: types_branch
> Reporter: Santhosh Srinivasan
> Assignee: Santhosh Srinivasan
> Fix For: types_branch
>
> Attachments: Error_handling_phase3_1.patch
>
>
> Phase three of the error handling feature will cover the backed including
> Hadoop specific error messages.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
