YanivKunda commented on code in PR #1573:
URL: https://github.com/apache/datafusion-comet/pull/1573#discussion_r2016335798
##########
spark/src/main/spark-3.5/org/apache/spark/sql/comet/shims/ShimCometScanExec.scala:
##########
@@ -55,15 +55,48 @@ trait ShimCometScanExec {
protected def isNeededForSchema(sparkSchema: StructType): Boolean = false
protected def getPartitionedFile(f: FileStatusWithMetadata, p:
PartitionDirectory): PartitionedFile =
- PartitionedFileUtil.getPartitionedFile(f, f.getPath, p.values)
+ try {
+ PartitionedFileUtil.getPartitionedFile(f, f.getPath, p.values)
+ } catch {
+ case _: NoSuchMethodError =>
Review Comment:
And keeping the current code paths? (i.e. compiled code for 3.5.5, else
reflection)
That is more explicit, but I'm not sure it brings any benefit.
The advantage of the try-catch idea that in case this param will go away in
3.5.6 (like it did in 4.0.0-preview1/2, while it's back in main...) this will
remain backward and forward-compatible.
--
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]
