[
https://issues.apache.org/jira/browse/FLINK-5303?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15804955#comment-15804955
]
ASF GitHub Bot commented on FLINK-5303:
---------------------------------------
Github user twalthr commented on a diff in the pull request:
https://github.com/apache/flink/pull/2976#discussion_r94965456
--- Diff:
flink-libraries/flink-table/src/main/scala/org/apache/flink/table/runtime/aggregate/AggregateUtil.scala
---
@@ -349,6 +354,19 @@ object AggregateUtil {
(groupingOffsetMapping, aggOffsetMapping)
}
+ private def getGroupingSetsMapping(outputType: RelDataType): Array[(Int,
Int)] = {
+ val fields = outputType.getFieldList
+ var mappingsBuffer = ArrayBuffer[(Int, Int)]()
+ for (i <- fields.indices) {
+ for (j <- fields.indices) {
+ if (fields(j).getName.equals("i$" + fields(i).getName)) {
--- End diff --
I know that it is unlikely but this could result in undesired behavior if
the table contains `i$test` and `test`. Isn't there a nicer way to determine
the grouping set mappings.
> Add CUBE/ROLLUP/GROUPING SETS operator in SQL
> ---------------------------------------------
>
> Key: FLINK-5303
> URL: https://issues.apache.org/jira/browse/FLINK-5303
> Project: Flink
> Issue Type: New Feature
> Components: Documentation, Table API & SQL
> Reporter: Alexander Chermenin
> Assignee: Alexander Chermenin
>
> Add support for such operators as CUBE, ROLLUP and GROUPING SETS in SQL.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
