[jira] [Updated] (CALCITE-5698) EXTRACT from INTERVAL partially does not follow the SQL standard

["ASF GitHub Bot (Jira)" <ji...@apache.org>]

     [ https://issues.apache.org/jira/browse/CALCITE-5698?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

ASF GitHub Bot updated CALCITE-5698:
------------------------------------
    Labels: pull-request-available  (was: )

> EXTRACT from INTERVAL partially does not follow the SQL standard
> ----------------------------------------------------------------
>
>                 Key: CALCITE-5698
>                 URL: https://issues.apache.org/jira/browse/CALCITE-5698
>             Project: Calcite
>          Issue Type: Bug
>          Components: core
>    Affects Versions: 1.34.0
>            Reporter: Evgeny Stanilovsky
>            Assignee: Evgeny Stanilovsky
>            Priority: Major
>              Labels: pull-request-available
>
> From SQL standard we can found:
> {noformat}
> ISO/IEC 9075-2:1999 (E)
> 6.17 <numeric value function>:
> If <extract field> is a <primary datetime field>, then the result is the value of the datetime
> field identified by that <primary datetime field> and has the same sign as the <extract
> source>.{noformat}
> other data bases are follow this rule, i.e. :
> {noformat}
> SELECT
>   EXTRACT (MONTH FROM INTERVAL '-1' MONTH)
> FROM
>   DUAL;{noformat}
> returns: *-1*
> and  **  the other one:
> {noformat}
> SELECT EXTRACT(MONTH FROM INTERVAL '-1 MONTHS'){noformat}
> return the same result,
> while calcite:
> {noformat}
> SELECT EXTRACT(MONTH FROM INTERVAL -1 MONTHS) AS extracted;{noformat}
> {noformat}
> < +---------+
> < |      -1 |
> < +---------+
> ---
> > +-----------+
> > |        11 |
> > +-----------+{noformat}



--
This message was sent by Atlassian Jira
(v8.20.10#820010)