xuechendi commented on a change in pull request #34396:
URL: https://github.com/apache/spark/pull/34396#discussion_r747962594
##########
File path: sql/core/src/main/scala/org/apache/spark/sql/execution/Columnar.scala
##########
@@ -458,6 +462,34 @@ case class RowToColumnarExec(child: SparkPlan) extends
RowToColumnarTransition {
// This avoids calling `schema` in the RDD closure, so that we don't need
to include the entire
// plan (this) in the closure.
val localSchema = this.schema
+ if (enableArrowColumnVector) {
+ val maxRecordsPerBatch = SQLConf.get.arrowMaxRecordsPerBatch
+ val timeZoneId = SQLConf.get.sessionLocalTimeZone
+ return child.execute().mapPartitionsInternal { rowIterator =>
+ val context = TaskContext.get()
+ val allocator = ArrowUtils.getDefaultAllocator
+ val bytesIterator = ArrowConverters
+ .toBatchIterator(rowIterator, localSchema, maxRecordsPerBatch,
timeZoneId, context)
Review comment:
yes, my question is why we need to add two extra physical plans?
If we make ArrowColumnVector as another backend of ColumnarBatch in
`RowsToColumnarExec`, it can also be used by scala version of `mapInArrow`,
doesn't it?
And another thing I am confused here is what is the `Arrow ` and `Columnar`
you are referring to in `ArrowToColumnarExec` and `ColumnarToArrowExec
`context? To me, arrow is columnar, so are your talking about adding two
Physical Plans to convert between `ArrowRecordBatch(type of RDD[Byte[Array]])`
and `ColumnarBatch(type of RDD[ColumnarBatch])`
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
