the-other-tim-brown commented on code in PR #13411:
URL: https://github.com/apache/hudi/pull/13411#discussion_r2143736659
##########
hudi-common/src/main/java/org/apache/hudi/metadata/HoodieMetadataPayload.java:
##########
@@ -376,19 +377,24 @@ public Option<IndexedRecord>
combineAndGetUpdateValue(IndexedRecord oldRecord, S
}
@Override
- public Option<IndexedRecord> getInsertValue(Schema schemaIgnored, Properties
propertiesIgnored) throws IOException {
+ public Option<IndexedRecord> getInsertValue(Schema schema, Properties
propertiesIgnored) throws IOException {
if (key == null || this.isDeletedRecord) {
return Option.empty();
}
HoodieMetadataRecord record = new HoodieMetadataRecord(key, type,
filesystemMetadata, bloomFilterMetadata,
columnStatMetadata, recordIndexMetadata, secondaryIndexMetadata);
- return Option.of(record);
+ if (schema == null ||
HoodieMetadataRecord.getClassSchema().equals(schema)) {
+ // If the schema is same or none is provided, we can return the record
directly
+ return Option.of(record);
+ } else {
+ return Option.of(rewriteRecord(record, schema));
Review Comment:
I dug into this a bit, the meta-fields are always present in the writer
schema when writing to the table so the outputs of the FGReader for the sake of
compaction are always expected to have the meta-fields. In order to avoid
unnecessary extra conversion, I am updating the code to directly create the
generic record so the merge output is in the expected 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]
