ramitg254 commented on code in PR #6259:
URL: https://github.com/apache/hive/pull/6259#discussion_r2715651062
##########
ql/src/java/org/apache/hadoop/hive/ql/parse/BaseSemanticAnalyzer.java:
##########
@@ -1345,7 +1348,14 @@ public TableSpec(Hive db, HiveConf conf, ASTNode ast,
boolean allowDynamicPartit
if (partHandle == null) {
// if partSpec doesn't exists in DB, return a delegate one
// and the actual partition is created in MoveTask
- partHandle = new Partition(tableHandle, partSpec, null);
+ try {
+ partHandle = tableHandle.hasNonNativePartitionSupport() ?
+ new DummyPartition(tableHandle,
Warehouse.makePartName(partSpec, false), partSpec) :
Review Comment:
by consistent passing I didn't mean this actually
I meant was similar to older behaviour irrespective of transform we were
adding the the provided partition spec directly to dummy partition without
applying the transform so should we keep that
or since you mentioned this inconsistency should we have a method in
IcebergTableUtil which build a transformed part spec for this purpose only so
that it can be used for both here and in getPartition to return that dummy
partition with that partition spec?
--
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]
