WweiL commented on code in PR #41558:
URL: https://github.com/apache/spark/pull/41558#discussion_r1227312489
##########
connector/connect/client/jvm/src/main/scala/org/apache/spark/sql/KeyValueGroupedDataset.scala:
##########
@@ -572,6 +726,101 @@ private class KeyValueGroupedDatasetImpl[K, V, IK, IV](
agg(aggregator)
}
+ override def mapGroupsWithState[S: Encoder, U: Encoder](
+ timeoutConf: GroupStateTimeout,
+ initialState: KeyValueGroupedDataset[K, S])(
+ func: (K, Iterator[V], GroupState[S]) => U): Dataset[U] = {
+ val nf = if (valueMapFunc == UdfUtils.identical()) {
+ func
+ } else {
+ UdfUtils.mapValuesAdaptor(func, valueMapFunc)
+ }
+ val outputEncoder = encoderFor[U]
+ sparkSession.newDataset[U](outputEncoder) { builder =>
+ builder.getFlatMapGroupsWithStateBuilder
+ .setInput(plan.getRoot)
+ .addAllGroupingExpressions(groupingExprs)
+ .setFunc(getUdf(nf, outputEncoder)(ivEncoder))
+ .setIsMapGroupsWithState(true)
+ .setOutputMode(OutputMode.Update.toString)
+ .setTimeoutConf(timeoutConf.toString)
+ .addAllInitialGroupingExpressions(initialState.getGroupingExpressions)
+ }
+ }
+
+ override def mapGroupsWithState[S: Encoder, U: Encoder](timeoutConf:
GroupStateTimeout)(
+ func: (K, Iterator[V], GroupState[S]) => U): Dataset[U] = {
+ val nf = if (valueMapFunc == UdfUtils.identical()) {
+ func
+ } else {
+ UdfUtils.mapValuesAdaptor(func, valueMapFunc)
+ }
+ val outputEncoder = encoderFor[U]
+ sparkSession.newDataset[U](outputEncoder) { builder =>
+ builder.getFlatMapGroupsWithStateBuilder
+ .setInput(plan.getRoot)
+ .addAllGroupingExpressions(groupingExprs)
+ .setFunc(getUdf(nf, outputEncoder)(ivEncoder))
+ .setIsMapGroupsWithState(true)
+ .setOutputMode(OutputMode.Update.toString)
+ .setTimeoutConf(timeoutConf.toString)
+ }
+ }
+
+ override def flatMapGroupsWithState[S: Encoder, U: Encoder](
+ outputMode: OutputMode,
+ timeoutConf: GroupStateTimeout,
+ initialState: KeyValueGroupedDataset[K, S])(
+ func: (K, Iterator[V], GroupState[S]) => Iterator[U]): Dataset[U] = {
+ if (outputMode != OutputMode.Append && outputMode != OutputMode.Update) {
+ throw new IllegalArgumentException("The output mode of function should
be append or update")
+ }
+ val nf = if (valueMapFunc == UdfUtils.identical()) {
+ func
+ } else {
+ UdfUtils.mapValuesAdaptor(func, valueMapFunc)
+ }
+ val outputEncoder = encoderFor[U]
+ sparkSession.newDataset[U](outputEncoder) { builder =>
+ builder.getFlatMapGroupsWithStateBuilder
+ .setInput(plan.getRoot)
+ .addAllGroupingExpressions(groupingExprs)
+ .setFunc(getUdf(nf, outputEncoder)(ivEncoder))
+ .setIsMapGroupsWithState(false)
+ .setOutputMode(outputMode.toString)
+ .setTimeoutConf(timeoutConf.toString)
+ .addAllInitialGroupingExpressions(initialState.getGroupingExpressions)
+ }
+ }
+
+ override def flatMapGroupsWithState[S: Encoder, U: Encoder](
+ outputMode: OutputMode,
+ timeoutConf: GroupStateTimeout)(
+ func: (K, Iterator[V], GroupState[S]) => Iterator[U]): Dataset[U] = {
+ if (outputMode != OutputMode.Append && outputMode != OutputMode.Update) {
+ throw new IllegalArgumentException("The output mode of function should
be append or update")
+ }
+ val nf = if (valueMapFunc == UdfUtils.identical()) {
+ func
+ } else {
+ UdfUtils.mapValuesAdaptor(func, valueMapFunc)
+ }
+ val outputEncoder = encoderFor[U]
+ sparkSession.newDataset[U](outputEncoder) { builder =>
+ builder.getFlatMapGroupsWithStateBuilder
+ .setInput(plan.getRoot)
+ .addAllGroupingExpressions(groupingExprs)
+ .setFunc(getUdf(nf, outputEncoder)(ivEncoder))
+ .setIsMapGroupsWithState(false)
+ .setOutputMode(outputMode.toString)
+ .setTimeoutConf(timeoutConf.toString)
+ }
+ }
+
+ override def getGroupingExpressions: java.util.List[proto.Expression] = {
Review Comment:
If my understand is correct, this function servers as a public getter of
`groupingExprs` so that `initialState.groupingExprs` can be called.
Can we follow the implementation here:
https://github.com/apache/spark/blob/db3f85b24498d4904134ed8f14b570fde6550060/connector/connect/client/jvm/src/main/scala/org/apache/spark/sql/KeyValueGroupedDataset.scala#L685-L703
To get rid of this function completely?
--
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]
