[CORE] Make LeafTransformSupport's getPartitions return Seq[Partition]

[Zouxxyy]

What changes are proposed in this pull request?

InputPartition is actually just a DSv2 interface API, Partition is common.
LeafTransformSupport's getPartitions should return Seq[Partition], for example:

FileSourceScanExecTransformer return Seq[FilePartition]
BatchScanExecTransformer return Seq[DataSourceRDDPartition]

Especially after Spark proposed Storage-Partitioned Join, a DataSourceRDDPartition may contain multi InputPartitions. See DataSourceRDD:

class DataSourceRDD(
    sc: SparkContext,
    @transient private val inputPartitions: Seq[Seq[InputPartition]],
    partitionReaderFactory: PartitionReaderFactory,
    columnarReads: Boolean,
    customMetrics: Map[String, SQLMetric])
  extends RDD[InternalRow](sc, Nil)

  override protected def getPartitions: Array[Partition] = {
    inputPartitions.zipWithIndex.map {
      case (inputPartitions, index) => new DataSourceRDDPartition(index, inputPartitions)
    }.toArray
  }
}

Due to this issue, some if-else codes were introduced previously, this PR cleans them up.

How was this patch tested?

[github-actions]

Run Gluten Clickhouse CI on x86

[github-actions]

Run Gluten Clickhouse CI on x86

[github-actions]

Run Gluten Clickhouse CI on x86

[github-actions]

Run Gluten Clickhouse CI on x86

[github-actions]

Run Gluten Clickhouse CI on x86

[github-actions]

Run Gluten Clickhouse CI on x86

[github-actions]

Run Gluten Clickhouse CI on x86

[github-actions]

Run Gluten Clickhouse CI on x86

[github-actions]

Run Gluten Clickhouse CI on x86

[github-actions]

Run Gluten Clickhouse CI on x86

[github-actions]

Run Gluten Clickhouse CI on x86

[github-actions]

Run Gluten Clickhouse CI on x86

[github-actions]

Run Gluten Clickhouse CI on x86

[github-actions]

Run Gluten Clickhouse CI on x86

[github-actions]

Run Gluten Clickhouse CI on x86

[github-actions]

Run Gluten Clickhouse CI on x86

[github-actions]

Run Gluten Clickhouse CI on x86

[github-actions]

Run Gluten Clickhouse CI on x86

[github-actions]

Run Gluten Clickhouse CI on x86

[github-actions]

Run Gluten Clickhouse CI on x86

[github-actions]

Run Gluten Clickhouse CI on x86

[github-actions]

Run Gluten Clickhouse CI on x86

[github-actions]

Run Gluten Clickhouse CI on x86

[Zouxxyy]

Run Gluten Clickhouse CI on x86

[zhztheplayer]

Thanks.

Should we also refactor GlutenWholeStageColumnarRDD? Does soft-affinity still work correctly after the change?

[Zouxxyy]

Should we also refactor GlutenWholeStageColumnarRDD?

For GlutenWholeStageColumnarRDD, I think these modifications below are sufficient. Do you think it needs to be modified in this PR as well? (Besieds. NativeFileScanColumnarRDD doesn't seem to be used. Should I delete it?)

case class FirstZippedPartitionsPartition(
    index: Int,
    inputPartition: Partition,
    inputColumnarRDDPartitions: Seq[Partition] = Seq.empty)
  extends Partition

class GlutenWholeStageColumnarRDD(
    @transient sc: SparkContext,
    @transient private val inputPartitions: Seq[Partition],

Does soft-affinity still work correctly after the change?

do you mean getPreferredLocations? I haven't modified its logic, so it still applies

e.g. in GlutenWholeStageColumnarRDD, It will do the correct cast.

  override def getPreferredLocations(split: Partition): Seq[String] = {
    castNativePartition(split)._1.preferredLocations()
  }

Or just like FileScanRDD in apache spark

  override protected def getPreferredLocations(split: RDDPartition): Seq[String] = {
    split.asInstanceOf[FilePartition].preferredLocations()
  }

[github-actions]

Run Gluten Clickhouse CI on x86

[zhztheplayer]

For GlutenWholeStageColumnarRDD, I think these modifications below are sufficient. Do you think it needs to be modified in this PR as well?

Yes and I saw you're working on this. Thanks.

Besieds. NativeFileScanColumnarRDD doesn't seem to be used. Should I delete it?

For sure. Thanks.

[github-actions]

Run Gluten Clickhouse CI on x86

[zhztheplayer]

Do we need to preserve a similar comment like this? Which will help user understand the .transpose usage here.

[github-actions]

Run Gluten Clickhouse CI on x86

[github-actions]

Run Gluten Clickhouse CI on x86

[github-actions]

Run Gluten Clickhouse CI on x86

[Zouxxyy]

@zhztheplayer I rebased and resolved the conflict, can you have a look, thanks

[zhztheplayer]

Thank you for the refactor. The code is much clearer now.