[
https://issues.apache.org/jira/browse/CASSANDRA-2124?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12999475#comment-12999475
]
Eric Evans commented on CASSANDRA-2124:
---------------------------------------
{quote}
Facing 1 issue with LongType decoder:
Query i am trying:
"UPDATE IndexedTable SET \"birthdate\" = 1000L WHERE KEY = 100L".
Was giving issue for long type. Looking further i can see that
{quote}
What kind of issue are you having? What is the comparator for this column
family (applied to column names)? What is the default validator, or the
specific one for column "birthdate" (applies to column values)?
{quote}
QueryProcessor->batchUpdate() method.
validateColumn is called with column.getKey().getByteBuffer() as a parameter.
Should it be
column.getValue().getByteBuffer()? Not sure but i was able to parse query
successfully with this change.
{quote}
This is looping over {{Map.Entry}} columns, so {{getKey()}} is returning the
column name, and {{getValue()}} the column value.
{quote}
Should we validate column value rather than it's name?
But still for selectquery
SELECT \"birthdate\" FROM IndexedTable WHERE KEY=100L. It is same error.
Again it is same thing, Should we validate column value rather than it's name?
QueryProcessor->getSlice method is doing the same.
Suggest, if i am trying with CQL in incorrect way.
{quote}
Both column names and values are validated against subclasses of
{{o.a.c.marshal.AbstractType}}, configured as comparators or validators
respectively.
> JDBC driver for CQL
> -------------------
>
> Key: CASSANDRA-2124
> URL: https://issues.apache.org/jira/browse/CASSANDRA-2124
> Project: Cassandra
> Issue Type: New Feature
> Components: API
> Reporter: Eric Evans
> Assignee: Vivek Mishra
> Priority: Minor
> Labels: cql
> Attachments: Cassandra-2124_v1.0, cassandra-0.7.1-2124_v2.0,
> cassandra-0.7.1-2124_v2.1, cassandra_generic_decoder.patch
>
>
> A simple connection class and corresponding pool was created for CQL as a
> part of CASSANDRA-1710, but a JDBC driver (either in addition to, or as a
> replacement for) would also be interesting.
--
This message is automatically generated by JIRA.
-
For more information on JIRA, see: http://www.atlassian.com/software/jira
