wesm commented on a change in pull request #7030: URL: https://github.com/apache/arrow/pull/7030#discussion_r416638207
########## File path: cpp/src/jni/dataset/proto/Types.proto ########## @@ -0,0 +1,149 @@ +// 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. + +syntax = "proto2"; +package types; + +option java_package = "org.apache.arrow.dataset"; Review comment: > As for filter, If I understand correctly it's OK to keep Java API but the JNI mapping for filters is considered fragile, right? I can remove the mapping anyway but when users read parquet files from Java they'll not be able to filter row groups to reduce I/O. Which is extremely important when low-selectivity filter is specified. I'm advising you to define some kind of abstraction layer so that the _details_ of the C++ API are not exposed in the Java API. So if there are changes in the C++ API, then refactoring can take place only at the JNI C++ code and no changes will be required in Java. ---------------------------------------------------------------- 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. For queries about this service, please contact Infrastructure at: [email protected]
