You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@sqoop.apache.org by "Jarek Jarcec Cecho (JIRA)" <ji...@apache.org> on 2014/11/11 01:41:33 UTC

[jira] [Commented] (SQOOP-1709) AbstractComplexType changes

    [ https://issues.apache.org/jira/browse/SQOOP-1709?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14205682#comment-14205682 ] 

Jarek Jarcec Cecho commented on SQOOP-1709:
-------------------------------------------

I've done investigation of the data types back in SQOOP-515. Having an array of floats or even array of arrays seems to be supported in some databases and hence it should be supported by our schema system as well. Map on the other hand doesn't seem to be supported on the relational databases that I've explored and I've added it because of Hive. The key in Hive can be any primitive type, e.g. it can be also a number, so String is not sufficient. It can't be an Array though. I think that for the purpose of powerness I've included the ability to have Map's key as complex type as well in case that we will add a database/storage that would have such support.

> AbstractComplexType changes
> ---------------------------
>
>                 Key: SQOOP-1709
>                 URL: https://issues.apache.org/jira/browse/SQOOP-1709
>             Project: Sqoop
>          Issue Type: Sub-task
>          Components: sqoop2-framework
>            Reporter: Veena Basavaraj
>            Assignee: Veena Basavaraj
>             Fix For: 1.99.5
>
>
> What is the use case for it to have a "key' as a Column type.
> This mean I could have the key as an Array / Map/ Flaoting point? when would this happen, can String not suffice for the key?
> I need details on this.
> Second, for good reasons, we should have a size field to support the collection size.
> Third, Make sure SchemaSerialization class is updated as well with these changes.



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