[
https://issues.apache.org/jira/browse/HIVE-11789?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14802713#comment-14802713
]
Jesus Camacho Rodriguez commented on HIVE-11789:
------------------------------------------------
[~ashutoshc], indeed that's the idea, to have better recognition so then we can
take advantage, or implement ourselves, further (generic) optimizations in
Calcite.
Further, if you look at HivePreFilteringRule, it seems clear that from the
engineering perspective, recognition of the functions using SqlKind when we go
into Calcite (and the inverse transformation) is better than recognizing the
function by comparing its annotation name in each rule independently.
> Better support for functions recognition in CBO
> -----------------------------------------------
>
> Key: HIVE-11789
> URL: https://issues.apache.org/jira/browse/HIVE-11789
> Project: Hive
> Issue Type: Improvement
> Components: CBO
> Affects Versions: 2.0.0
> Reporter: Jesus Camacho Rodriguez
> Assignee: Jesus Camacho Rodriguez
> Attachments: HIVE-11789.01.patch, HIVE-11789.patch
>
>
> Including IN, BETWEEN, and STRUCT (multicolumn). In particular, STRUCT is
> represented as ROW in Calcite.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
