[jira] [Comment Edited] (CALCITE-6886) RuleMatchVisualizer should merge sets when the planner detects their equivalence

["Thomas Rebele (Jira)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/CALCITE-6886?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17934497#comment-17934497 ] 

Thomas Rebele edited comment on CALCITE-6886 at 3/14/25 11:10 AM:
------------------------------------------------------------------

In the example, #10-PhysLeafRel appears out of the blue in the FINAL step. Good opportunity to check why this is the case.

Another minor thing to fix: if there's a ?step=33 argument, nothing is viewed because of an error: data.steps[prevStepIndex] is undefined.


was (Author: thomas.rebele):
In the example, #10-PhysLeafRel appears out of the blue in the FINAL step. Good opportunity to check why this is the case.

> RuleMatchVisualizer should merge sets when the planner detects their equivalence
> --------------------------------------------------------------------------------
>
>                 Key: CALCITE-6886
>                 URL: https://issues.apache.org/jira/browse/CALCITE-6886
>             Project: Calcite
>          Issue Type: Improvement
>            Reporter: Thomas Rebele
>            Priority: Minor
>         Attachments: CALCITE-6886.png
>
>
> If the planner merges two sets when they have been determined to be equivalent, the RuleMatchVisualizer does not merge them in the visualization. Instead, it keeps the second set with the subsets but no RelNode inside.
> To reproduce the issue, apply the RuleMatchVisualizer to VolcanoPlannerTest.testSetMergeMatchSubsetRule(). The set-2 should be merged into set-0, and the subset#11-set#2 into subset#12-set#0:
> !CALCITE-6886.png!
>  



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