linliu-code commented on code in PR #9593:
URL: https://github.com/apache/hudi/pull/9593#discussion_r1326462320
##########
hudi-spark-datasource/hudi-spark-common/src/main/java/org/apache/hudi/HoodieSparkRecordMerger.java:
##########
@@ -38,44 +37,87 @@ public String getMergingStrategy() {
}
@Override
- public Option<Pair<HoodieRecord, Schema>> merge(HoodieRecord older, Schema
oldSchema, HoodieRecord newer, Schema newSchema, TypedProperties props) throws
IOException {
- ValidationUtils.checkArgument(older.getRecordType() ==
HoodieRecordType.SPARK);
- ValidationUtils.checkArgument(newer.getRecordType() ==
HoodieRecordType.SPARK);
+ public Option<Pair<HoodieRecord, Schema>> merge(Option<HoodieRecord> older,
+ Schema oldSchema,
+ Option<HoodieRecord> newer,
+ Schema newSchema,
+ TypedProperties props)
throws IOException {
+ boolean isValidNew = isValid(newer, newSchema, props);
+ boolean isValidOld = isValid(older, oldSchema, props);
- if (newer instanceof HoodieSparkRecord) {
- HoodieSparkRecord newSparkRecord = (HoodieSparkRecord) newer;
- if (newSparkRecord.isDeleted()) {
- // Delete record
- return Option.empty();
- }
+ if (!isValidOld && !isValidNew) {
+ return handleTrivialCase(older, oldSchema, newer, newSchema, props);
+ } else if (isValidOld && !isValidNew) {
+ return handleDeleteCase(older, oldSchema, newer, newSchema, props);
+ } else if (!isValidOld) {
+ return handleInsertCase(older, oldSchema, newer, newSchema, props);
} else {
- if (newer.getData() == null) {
- // Delete record
- return Option.empty();
- }
- }
-
- if (older instanceof HoodieSparkRecord) {
- HoodieSparkRecord oldSparkRecord = (HoodieSparkRecord) older;
- if (oldSparkRecord.isDeleted()) {
- // use natural order for delete record
- return Option.of(Pair.of(newer, newSchema));
- }
- } else {
- if (older.getData() == null) {
- // use natural order for delete record
- return Option.of(Pair.of(newer, newSchema));
- }
- }
- if (older.getOrderingValue(oldSchema,
props).compareTo(newer.getOrderingValue(newSchema, props)) > 0) {
- return Option.of(Pair.of(older, oldSchema));
- } else {
- return Option.of(Pair.of(newer, newSchema));
+ return handleUpdateCase(older, oldSchema, newer, newSchema, props);
}
}
@Override
public HoodieRecordType getRecordType() {
return HoodieRecordType.SPARK;
}
+
+ private boolean isValid(Option<HoodieRecord> record, Schema schema,
TypedProperties props) throws IOException {
+ return record.isPresent()
+ && !(record.get() instanceof HoodieEmptyRecord)
+ && !(record.get().isDelete(schema, props))
+ && record.get().getRecordType() == HoodieRecordType.SPARK;
+ }
+
+ /**
+ * Handle trivial case.
+ */
+ Option<Pair<HoodieRecord, Schema>> handleTrivialCase(
+ Option<HoodieRecord> olderIgnored,
+ Schema oldSchemaIgnored,
+ Option<HoodieRecord> newerIgnored,
+ Schema newSchemaIgnored,
+ TypedProperties propsIgnored) throws IOException {
+ return Option.empty();
+ }
+
+ /**
+ * Handle delete case.
+ */
+ Option<Pair<HoodieRecord, Schema>> handleDeleteCase(
+ Option<HoodieRecord> olderIgnored,
+ Schema oldSchemaIgnored,
+ Option<HoodieRecord> newerIgnored,
+ Schema newSchemaIgnored,
+ TypedProperties propsIgnored) throws IOException {
+ return Option.empty();
Review Comment:
Either one should work, but checking `option.isPresent()` should be more
efficient than `instanceOf`. Wdyt?
--
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]