[
https://issues.apache.org/jira/browse/CALCITE-4673?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17372501#comment-17372501
]
Stanilovsky Evgeny commented on CALCITE-4673:
---------------------------------------------
[~julianhyde] i check, different columns will generate :
{noformat}
LogicalTableFunctionScan(invocation=[DEDUP($cor0.EMPNO, $cor0.DEPTNO)],
rowType=[RecordType(VARCHAR(1024) NAME)])
{noformat}
this issue occurs for example in our engine with custom udf :
{noformat}
SELECT t.id, (SELECT x FROM TABLE(system_range(t.id, t.id))) FROM random_tbl t;
{noformat}
will raise two different correlates, which breaks further execution. I try to
mirroring our problem into pure calcite code.
> Correlate variables deduplication is not triggered when custom table function
> presented.
> ----------------------------------------------------------------------------------------
>
> Key: CALCITE-4673
> URL: https://issues.apache.org/jira/browse/CALCITE-4673
> Project: Calcite
> Issue Type: Bug
> Components: core
> Affects Versions: 1.27.0
> Reporter: Stanilovsky Evgeny
> Assignee: Stanilovsky Evgeny
> Priority: Major
> Labels: pull-request-available
> Time Spent: 10m
> Remaining Estimate: 0h
>
> request like:
> {noformat}
> select e.deptno, (select * from lateral table(DEDUP(e.deptno, e.deptno)))
> from emp e;
> {noformat}
> shows 2 different correlates, while it only one.
> {noformat}
> LogicalProject(DEPTNO=[$7], EXPR$1=[$SCALAR_QUERY({
> LogicalProject(NAME=[$0])
> LogicalTableFunctionScan(invocation=[DEDUP($cor0.DEPTNO, $cor1.DEPTNO)],
> rowType=[RecordType(VARCHAR(1024) NAME)])
> })])
> LogicalTableScan(table=[[CATALOG, SALES, EMP]])
> {noformat}
> [dev-list
> discussion|https://lists.apache.org/thread.html/re20d580a979585ae946d61df80a345c70ae2b5a92577a8b5db1e2c50%40<dev.calcite.apache.org>]
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
