Github user watermen commented on a diff in the pull request:
https://github.com/apache/spark/pull/13321#discussion_r64863306
--- Diff:
sql/core/src/main/scala/org/apache/spark/sql/execution/ExistingRDD.scala ---
@@ -336,9 +336,11 @@ private[sql] object DataSourceScanExec {
def create(
output: Seq[Attribute],
rdd: RDD[InternalRow],
- relation: BaseRelation,
+ logicalRelation: LogicalRelation,
metadata: Map[String, String] = Map.empty,
metastoreTableIdentifier: Option[TableIdentifier] = None):
DataSourceScanExec = {
+ val relation: BaseRelation = logicalRelation.relation
+
val outputPartitioning = {
val bucketSpec = relation match {
--- End diff --
Do you mean?
```scala
def toAttribute(colName: String): Attribute = output.find(_.name ==
colName).get
bucketSpec.map { spec =>
val numBuckets = spec.numBuckets
if (spec.bucketColumnNames.forall(colName => output.exists(_.name ==
colName))) {
val bucketColumns = spec.bucketColumnNames.map(toAttribute)
HashPartitioning(bucketColumns, numBuckets)
} else {
UnknownPartitioning(0)
}
}.getOrElse {
UnknownPartitioning(0)
}
```
But the sql below also can't be benefited from bucketing but it's satisfy
all conditions(the conf is enabled, and the relation is bucketed, and the
output contains all bucketing columns).
```sql
SELECT i, j, k FROM bucketed_table;
```
So we need to use `groupbyColums` or `joinColums` instead of `output`?
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at [email protected] or file a JIRA ticket
with INFRA.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
