haiyangsun-db commented on code in PR #49260:
URL: https://github.com/apache/spark/pull/49260#discussion_r1896557761
##########
connector/connect/client/jvm/src/main/scala/org/apache/spark/sql/KeyValueGroupedDataset.scala:
##########
@@ -502,18 +502,20 @@ private class KeyValueGroupedDatasetImpl[K, V, IK, IV](
}
val outputEncoder = agnosticEncoderFor[U]
+ val stateEncoder = agnosticEncoderFor[S]
val nf = UDFAdaptors.flatMapGroupsWithStateWithMappedValues(func,
valueMapFunc)
sparkSession.newDataset[U](outputEncoder) { builder =>
val groupMapBuilder = builder.getGroupMapBuilder
groupMapBuilder
.setInput(plan.getRoot)
.addAllGroupingExpressions(groupingExprs)
- .setFunc(getUdf(nf, outputEncoder)(ivEncoder))
+ .setFunc(getUdf(nf, outputEncoder, stateEncoder)(ivEncoder))
.setIsMapGroupsWithState(isMapGroupWithState)
.setOutputMode(if (outputMode.isEmpty) OutputMode.Update.toString
else outputMode.get.toString)
.setTimeoutConf(timeoutConf.toString)
+ .setStateSchema(stateEncoder.schema.json)
Review Comment:
A better way to pass a schema (StructType) in spark connect than using json,
see:
https://github.com/apache/spark/blob/master/connector/connect/client/jvm/src/main/scala/org/apache/spark/sql/Dataset.scala#L197
.setStateSchema(DataTypeProtoConverter.toConnectProtoType(stateEncoder.schema))
--
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]
