[jira] [Created] (IMPALA-10452) CREATE Iceberg tables with old PARTITIONED BY syntax

["Zoltán Borók-Nagy (Jira)" <ji...@apache.org>]

Zoltán Borók-Nagy created IMPALA-10452:
------------------------------------------

             Summary: CREATE Iceberg tables with old PARTITIONED BY syntax
                 Key: IMPALA-10452
                 URL: https://issues.apache.org/jira/browse/IMPALA-10452
             Project: IMPALA
          Issue Type: Test
            Reporter: Zoltán Borók-Nagy


It's convenient for users to create Iceberg tables with the old syntax.

It's also easier to migrate existing workloads to Iceberg because the SQL scripts that create the table definitions don't need to change that much.

So users should be able to write the following:

 
{noformat}
CREATE TABLE ice_t (i int)
PARTITIONED BY (p int)
STORED AS ICEBERG;
{noformat}
Which should be equivalent to this:

 
{noformat}
CREATE TABLE ice_t (i int, p int)
PARTITION BY SPEC (p IDENTITY)
STORED AS ICEBERG;
{noformat}
 

Please note that the old-style CREATE TABLE creates IDENTITY-partitioned tables. For other partition transforms the users must use the new, more generic syntax.

Hive also supports the PARTITIONED BY syntax, see [https://github.com/apache/iceberg/pull/1612]



--
This message was sent by Atlassian Jira
(v8.3.4#803005)