morrySnow commented on code in PR #65569:
URL: https://github.com/apache/doris/pull/65569#discussion_r3602103348
##########
fe/fe-core/src/main/java/org/apache/doris/nereids/trees/plans/physical/PhysicalLazyMaterializeFileScan.java:
##########
@@ -55,16 +68,25 @@ public List<String> getQualifier() {
@Override
public List<Slot> computeOutput() {
Review Comment:
Same caching-removal pattern as PhysicalLazyMaterializeTVFScan. The old code
cached computeOutput() with an output field guard. The new code recomputes on
every call. Consider restoring the cached output for consistency and to avoid
repeated list allocations during plan traversal.
##########
fe/fe-core/src/main/java/org/apache/doris/nereids/trees/plans/visitor/PlanVisitor.java:
##########
@@ -301,15 +301,15 @@ public R
visitPhysicalLazyMaterialize(PhysicalLazyMaterialize<? extends Plan> ma
}
public R
visitPhysicalLazyMaterializeFileScan(PhysicalLazyMaterializeFileScan scan, C
context) {
- return visit(scan, context);
+ return visitPhysicalFileScan(scan, context);
}
public R
visitPhysicalLazyMaterializeTVFScan(PhysicalLazyMaterializeTVFScan scan, C
context) {
- return visit(scan, context);
+ return visitPhysicalTVFRelation(scan, context);
}
public R
visitPhysicalLazyMaterializeOlapScan(PhysicalLazyMaterializeOlapScan scan, C
context) {
Review Comment:
This delegation fix (visitPhysicalLazyMaterializeOlapScan ->
visitPhysicalOlapScan, and same for FileScan/TVFScan) is a good cleanup. It
eliminates the need for every visitor to add boilerplate overrides for the
lazy-scan wrapper nodes. This is what enables the removal of the
visitPhysicalLazyMaterializeOlapScan override from RuntimeFilterGenerator and
simplifies TopnFilterPushDownVisitor.
--
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]
