kkraus14 commented on a change in pull request #10979:
URL: https://github.com/apache/arrow/pull/10979#discussion_r695347914



##########
File path: format/IRFunction.fbs
##########
@@ -0,0 +1,68 @@
+// Licensed to the Apache Software Foundation (ASF) under one
+// or more contributor license agreements.  See the NOTICE file
+// distributed with this work for additional information
+// regarding copyright ownership.  The ASF licenses this file
+// to you under the Apache License, Version 2.0 (the
+// "License"); you may not use this file except in compliance
+// with the License.  You may obtain a copy of the License at
+//
+//   http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing,
+// software distributed under the License is distributed on an
+// "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
+// KIND, either express or implied.  See the License for the
+// specific language governing permissions and limitations
+// under the License.
+
+include "Schema.fbs";
+
+namespace org.apache.arrow.ir.flatbuf;
+
+// A unique identifier for a particular function definition. 
+table FunctionId {
+  // the function description identifier.
+  id: uint32;
+  
+  // The origin of the function definition. Should be mapped to a domain such 
as "org.apache.arrow"
+  // The github definition paths for definitions are defined in 
+  // github.com/apache/arrow/format/functions.txt. Org 0 == Apache Arrow 
canonical definitions.
+  // Organizations < 1B are canonical organizations. For private functions, use
+  // Org id > 1B

Review comment:
       I'm concerned about having namespaced functions in the logical plan that 
aren't user defined functions.
   
   For example, there's already functions and engines implemented in both C++ 
and Rust. If someone builds a logical plan from say a Python or R library, I 
imagine that logical plan should be able to be executed by either the C++ or 
Rust engine and that engine could choose which function implementation they 
should map to when building their physical plan.
   
   I.E. say there's some arbitrary function that is standardized like 
`functionA` and there's both a C++ and Rust implementation of it. Would they 
have the same function IDs in different orgs or would they have the same org 
with different function IDs? How would an IR producer allow arbitrarily 
targeting either implementation?




-- 
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]


Reply via email to