andygrove commented on code in PR #1793:
URL: https://github.com/apache/datafusion-comet/pull/1793#discussion_r2109546495
##########
native/core/src/execution/planner.rs:
##########
@@ -956,6 +956,12 @@ impl PhysicalPlanner {
self.create_expr(filter.predicate.as_ref().unwrap(),
child.schema())?;
let filter: Arc<dyn ExecutionPlan> = if
filter.use_datafusion_filter {
+ assert_eq!(
+ scans.len(),
+ 0,
+ "Must not use ScanExec as it may requires a copy"
+ );
+
Review Comment:
My understanding is that we only reuse buffers when using `native_comet` to
read the Parquet files. If `native_iceberg_compat` is used then we will still
have a `ScanExec` reading from JVM but it is safe to use DataFusion's
FilterExec in this case.
@parthchandra would you agree with this?
--
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]
