[
https://issues.apache.org/jira/browse/CASSANDRA-9186?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14547950#comment-14547950
]
Aleksey Yeschenko commented on CASSANDRA-9186:
----------------------------------------------
Aside from CASSANDRA-6717 comments, should we really make a frozen/non-frozen
distinction at UDF level at all?
We don't do that in the native protocol.
> AbstractType vs. CQL3Type loses frozen keyword
> ----------------------------------------------
>
> Key: CASSANDRA-9186
> URL: https://issues.apache.org/jira/browse/CASSANDRA-9186
> Project: Cassandra
> Issue Type: Bug
> Components: Core
> Reporter: Robert Stupp
> Assignee: Robert Stupp
> Priority: Minor
> Fix For: 3.x
>
>
> {code}
> create keyspace functionmetadatatest with replication =
> {'class':'SimpleStrategy', 'replication_factor':1};
> use functionmetadatatest ;
> CREATE TYPE udtx (x int);
> CREATE FUNCTION u_frozen(u frozen<udtx>)RETURNS int LANGUAGE java AS 'return
> new Integer(0);';
> SELECT function_name, signature, argument_types FROM system.schema_functions
> WHERE keyspace_name='functionmetadatatest';
> function_name | signature | argument_types
> ---------------+------------------+--------------------------------------------------------------------------------------------------------------------------
> u_frozen | ['udtx'] |
> ['org.apache.cassandra.db.marshal.UserType(functionmetadatatest,75647478,78:org.apache.cassandra.db.marshal.Int32Type)']
> {code}
> Problem is that {{UserType}} and {{TupleType}} do not have a _frozen_ (or
> _isMultiCell_) attribute. A conversion from type to cql3type therefore loses
> the _frozen_ status.
> EDIT: The issue has been discovered by [~aholmber]
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
