[
https://issues.apache.org/jira/browse/PIG-1256?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12837951#action_12837951
]
Xuefu Zhang commented on PIG-1256:
----------------------------------
1. All test case failure seemed due to env issues. Manually running the failed
cases succeeds.
2. In this patch, schemas for a collection field such as f:collection(f1:int,
f2:string) and f:collection(record(f1:int, f2:string)) are treated as the same
by both Zebra and by PIG when they are converted to Pig ResourceSchema and
later in Pig to Pig schema. Previously when Zebra schema is directly converted
to Pig schema, the second case has no second level access in Pig script. Now
both the cases has second level access.
> [Zebra] Bag field should always contain a tuple type as the field schema in
> ResourceSchema object converted from Zebra Schema
> -----------------------------------------------------------------------------------------------------------------------------
>
> Key: PIG-1256
> URL: https://issues.apache.org/jira/browse/PIG-1256
> Project: Pig
> Issue Type: Bug
> Components: impl
> Affects Versions: 0.7.0
> Reporter: Xuefu Zhang
> Assignee: Xuefu Zhang
> Priority: Minor
> Fix For: 0.7.0
>
> Attachments: patch.0223
>
>
> Pig now requires that schema converted from Zebra Schema contains Tuple type
> as field schema in .7 release. Zebra needs to take care of all cases where
> Record is not explicitly specified.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
