sohami commented on a change in pull request #1417: DRILL-6645: Transform TopN
in Lateral Unnest pipeline to Sort and Limit.
URL: https://github.com/apache/drill/pull/1417#discussion_r207583353
##########
File path:
exec/java-exec/src/main/java/org/apache/drill/exec/planner/physical/Prel.java
##########
@@ -56,7 +56,13 @@ public boolean
useAbstractConvertersForConversion(RelTraitSet fromTraits,
SelectionVectorMode getEncoding();
boolean needsFinalColumnReordering();
- default Prel addImplicitRowIDCol(List<RelNode> children) {
+ /**
+ * If the operator is in Lateral/Unnest pipeline, then it generates a new
operator which knows how to process
+ * the rows accordingly during execution.
+ * eg: TopNPrel -> SortPrel and LimitPrel
+ * Other operators like FilterPrel, ProjectPrel etc will add an implicit row
id to the output.
+ */
+ default Prel getRelForLateralUnnestPipeline(List<RelNode> children) {
throw new UnsupportedOperationException("Adding Implicit RowID column is
not supported for " +
Review comment:
This function name is changed from `addImplicitRowIdCol` to
`getRelForLateralUnnestPipeline` but from comments it looks like it does both
the functionality. Based on my understanding for all operators except
`TopNPrel` it's actually adding the `implicitRowIdCol` information, but since
topN will not be supported inside subquery and needs to be replaced by
PartitionLimit + Sort we are transforming it inside this function. And the new
limit and Sort will have the information about the implicitRowIdCol. If this is
true, then we should keep the name same as before and in TopN as you have done
calling separate private function `transformTopNToSortAndLimit` is fine.
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
[email protected]
With regards,
Apache Git Services
