[ 
https://issues.apache.org/jira/browse/PIG-927?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
 ]

Daniel Dai updated PIG-927:
---------------------------

    Attachment: PIG-927-1.patch

In the patch, we follow SQL behavior. When we join on more than one key (it is 
a tuple key in Pig), as long as one of keys is null, we do not merge them. Eg: 
we do not merge below tuple pair:

(1, 2, null) vs (1, 2, null)

> null should be handled consistently in Join
> -------------------------------------------
>
>                 Key: PIG-927
>                 URL: https://issues.apache.org/jira/browse/PIG-927
>             Project: Pig
>          Issue Type: Bug
>    Affects Versions: 0.3.0
>            Reporter: Pradeep Kamath
>         Attachments: PIG-927-1.patch
>
>
> Currenlty Pig mostly follows SQL semantics for handling null. However there 
> are certain cases where pig may need to handle nulls correctly. One example 
> is the join - joins on single keys results in null keys not matching to 
> produce an output. However if the join is on >1 keys, in the key tuple, if 
> one of the values is null, it still matches with another key tuple which has 
> a null for that value. We need to decide the right semantics here. 

-- 
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.

Reply via email to