[
https://issues.apache.org/jira/browse/CASSANDRA-1618?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13009326#comment-13009326
]
Brandon Williams commented on CASSANDRA-1618:
---------------------------------------------
We'd have to use \u notation. The problem is, we can't know which comparator
to use during the parse/setup phase of the job. We could ease this somewhat by
providing UDFs that would generate the correct \u notation, basically exposing
comparator.fromString and then converting that to \u notation for pig.
Alternately, we could cheat in the loader and describe the ring ourselves, but
that would be going outside of CFIF/CFRR.
> Allow specifying a slice predicate for pig queries to filter out columns
> ------------------------------------------------------------------------
>
> Key: CASSANDRA-1618
> URL: https://issues.apache.org/jira/browse/CASSANDRA-1618
> Project: Cassandra
> Issue Type: Improvement
> Components: Contrib, Hadoop
> Reporter: Jeremy Hanna
> Assignee: Brandon Williams
> Fix For: 0.7.5
>
> Attachments: 1618.txt
>
>
> Currently the Cassandra/Pig loadfunc hardcodes the slice predicate to get all
> columns in a particular row. It would be nice to allow the user to configure
> the slice predicate.
--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira
