codope commented on code in PR #12720:
URL: https://github.com/apache/hudi/pull/12720#discussion_r1932476018
##########
hudi-spark-datasource/hudi-spark/src/main/scala/org/apache/spark/sql/hudi/command/payload/ExpressionPayload.scala:
##########
@@ -164,6 +165,47 @@ class ExpressionPayload(@transient record: GenericRecord,
}
}
+ private def doRecordMerge(incomingRecord: GenericRecord,
+ existingRecord: IndexedRecord,
+ schema: Schema,
+ properties: Properties): HOption[IndexedRecord] = {
+ val tablePayloadClass =
properties.getProperty(PAYLOAD_ORIGINAL_AVRO_PAYLOAD)
+ if
(tablePayloadClass.equals(classOf[OverwriteWithLatestAvroPayload].getName)) {
+ HOption.of(incomingRecord)
+ } else if
(tablePayloadClass.equals(classOf[DefaultHoodieRecordPayload].getName)) {
+ if
(needUpdatingPersistedRecord(existingRecord.asInstanceOf[GenericRecord],
incomingRecord, properties)) {
+ HOption.of(incomingRecord)
+ } else {
+ HOption.of(existingRecord)
+ }
+ } else {
+ val orderingField = ConfigUtils.getOrderingField(properties)
+ if (StringUtils.isNullOrEmpty(orderingField)) {
+ HOption.of(incomingRecord)
+ } else {
+ val consistentLogicalTimestampEnabled =
properties.getProperty(KeyGeneratorOptions.KEYGENERATOR_CONSISTENT_LOGICAL_TIMESTAMP_ENABLED.key,
+
KeyGeneratorOptions.KEYGENERATOR_CONSISTENT_LOGICAL_TIMESTAMP_ENABLED.defaultValue).toBoolean
+ val incomingRecordPayload =
HoodieRecordUtils.loadPayload(tablePayloadClass, incomingRecord,
+ HoodieAvroUtils.getNestedFieldVal(incomingRecord, orderingField,
true, consistentLogicalTimestampEnabled)
+ .asInstanceOf[Comparable[_]]).asInstanceOf[HoodieRecordPayload[_
<: HoodieRecordPayload[_]]]
Review Comment:
for CUSTOM merge mode, we need to load the payload class whose construction
is based on incoming record, so i don't think we can do this upfront with only
record payload API. Perhaps, we can use record merge API but it's a larger
refactoring.
--
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]
