[
https://issues.apache.org/jira/browse/CASSANDRA-7351?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14034289#comment-14034289
]
Aleksey Yeschenko commented on CASSANDRA-7351:
----------------------------------------------
Looks correct to me. That said, looking at the bigger picture, we should
refactor BatchStatement a bit (this does contradict my initial remarks, sorry).
1. Should move all the validation from Parsed#prepare() to
BatchStatement#validate() - including "Batch with conditions cannot span
multiple tables"
2. Should remove, really, not update, the counter validation logic from
BatchMessage, and rely on BS#validate() instead - it will get called by
processBatch() down the line. No need to duplicate this logic in two places.
> Make BEGIN COUNTER BATCH syntax optional
> ----------------------------------------
>
> Key: CASSANDRA-7351
> URL: https://issues.apache.org/jira/browse/CASSANDRA-7351
> Project: Cassandra
> Issue Type: Improvement
> Reporter: T Jake Luciani
> Assignee: T Jake Luciani
> Priority: Trivial
> Labels: cql3
> Fix For: 2.1.0
>
> Attachments: 7351-rebase.txt, 7351.txt, 7351v2.txt
>
>
> AFAIK there is no need to supply COUNTER in Batch statements.
> The server should just throw an syntax error if you try to mix counter and
> non counter statements in the same batch.
> What value does this keyword add?
--
This message was sent by Atlassian JIRA
(v6.2#6252)
