parthchandra commented on code in PR #620:
URL: https://github.com/apache/datafusion-comet/pull/620#discussion_r1663091296
##########
spark/src/main/scala/org/apache/comet/serde/QueryPlanSerde.scala:
##########
@@ -2141,6 +2141,25 @@ object QueryPlanSerde extends Logging with
ShimQueryPlanSerde with CometExprShim
scalarExprToProtoWithReturnType(algorithm, StringType, childExpr)
}
+ case struct @ CreateNamedStruct(_) =>
+ val valExprs = struct.valExprs.map(exprToProto(_, inputs, binding))
+ val dataType = serializeDataType(struct.dataType)
+
+ if (valExprs.forall(_.isDefined) && dataType.isDefined) {
+ val structBuilder = ExprOuterClass.CreateNamedStruct.newBuilder()
+ structBuilder.addAllValues(valExprs.map(_.get).asJava)
+ structBuilder.setDatatype(dataType.get)
+
+ Some(
+ ExprOuterClass.Expr
+ .newBuilder()
+ .setCreateNamedStruct(structBuilder)
+ .build())
+ } else {
+ withInfo(expr, struct.valExprs: _*)
Review Comment:
> I think we need to specify an info string here otherwise the user will not
see the reason for falling back to Spark (perhaps @parthchandra can confirm).
The user will get the reason from the child nodes. We could add the
additional message as @andygrove has suggested, but it is not strictly
necessary.
--
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]
