[jira] [Comment Edited] (CASSANDRA-9390) AbstractColumnFamilyInputFormat doesn't work with mixed case keyspaces

["Mike Adamson (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/CASSANDRA-9390?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14545144#comment-14545144 ] 

Mike Adamson edited comment on CASSANDRA-9390 at 5/15/15 10:09 AM:
-------------------------------------------------------------------

[~philipthompson] The problem is that we either go with all the hadoop code quoting these identifiers or none of it. At the moment most of the code does quote the identifiers so we would need to remove it if we don't want to do it here. I was discussing this with [~jjordan] and his view was that we shouldn't be quoting. If we don't quote then it is going to mean changes for any customers using the hadoop formats. If we do want to remove the quoting I suggest we do it on another ticket.

EDIT
This is the only instance of missing quotes I could find so far. If I find any others I'll raise separate jiras.


was (Author: mike_tr_adamson):
[~philipthompson] The problem is that we either go with all the hadoop code quoting these identifiers or none of it. At the moment most of the code does quote the identifiers so we would need to remove it if we don't want to do it here. I was discussing this with [~jjordan] and his view was that we shouldn't be quoting. If we don't quote then it is going to mean changes for any customers using the hadoop formats. If we do want to remove the quoting I suggest we do it on another ticket.

> AbstractColumnFamilyInputFormat doesn't work with mixed case keyspaces
> ----------------------------------------------------------------------
>
>                 Key: CASSANDRA-9390
>                 URL: https://issues.apache.org/jira/browse/CASSANDRA-9390
>             Project: Cassandra
>          Issue Type: Bug
>          Components: Hadoop
>            Reporter: Mike Adamson
>            Assignee: Philip Thompson
>             Fix For: 2.2.x
>
>         Attachments: 9390-1.txt
>
>
> The AbstractColumnFamilyInputFormat isn't quoting the keyspace before calling metadata.getReplicas:
> {noformat}
>         for (TokenRange tokenRange : metadata.getTokenRanges())
>             map.put(tokenRange, metadata.getReplicas(keyspace, tokenRange));
>         return map;
> {noformat}
> This is inconsistent with other areas of the hadoop code and causes the inputs using mixed case keyspaces to fail.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)