[ 
https://issues.apache.org/jira/browse/CASSANDRA-7461?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15310543#comment-15310543
 ] 

Benjamin Lerer commented on CASSANDRA-7461:
-------------------------------------------

{quote}
Do you think there is that much difference between those to justify spitting it?
{quote}

I asked myself that question. On the java side I do not think it is the case 
but each type of operators need to be tested and documented. Some changes needs 
also to be done to {{CQLSH}} for auto completion. In the end I came to the 
conclusion that splitting the ticket will reduce the risk of forgetting stuff 
and make the review process easier but I might be wrong.

{quote}
it's that part that I would split in a separate ticket (since again, it's not 
specific to operators, even if operators are a good motivation for the 
improvement). 
{quote}

It is definetly something that we should do.

> operator functionality in CQL
> -----------------------------
>
>                 Key: CASSANDRA-7461
>                 URL: https://issues.apache.org/jira/browse/CASSANDRA-7461
>             Project: Cassandra
>          Issue Type: New Feature
>          Components: CQL
>            Reporter: Robert Stupp
>            Assignee: Benjamin Lerer
>              Labels: cql
>
> Intention: Allow operators in CQL
> Operators could be decimal arithmetics {{+ - * /}} or boolen arithmetics {{| 
> & !}} or string 'arithmetics' {{+}}
> {{SELECT tab.label + ' = ' + tab.value FROM foo.tab}}
> {{SELECT * FROM tab WHERE tab.label + ' = ' + tab.value = 'foo = bar'}}
> as well as
> {{CREATE INDEX idx ON tab ( tab.tabel + '=' + tab.value )}}
> or
> {{CREATE INDEX idx ON tab (label) WHERE contains(tab.tabel, 
> 'very-important-key')}}
> Operators could be mapped to UDFs like this:
> {{+}} mapped to UDF {{cstarstd::oper_plus(...)}}
> {{-}} mapped to UDF {{cstarstd::oper_minus(...)}}
> or handled directly via {{Cql.g}} in 'special' code



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

Reply via email to