pvary commented on code in PR #11513: URL: https://github.com/apache/iceberg/pull/11513#discussion_r1877424024
########## core/src/main/java/org/apache/iceberg/actions/FileRewriteGroup.java: ########## @@ -0,0 +1,93 @@ +/* + * 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. + */ +package org.apache.iceberg.actions; + +import java.util.Comparator; +import java.util.List; +import org.apache.iceberg.ContentFile; +import org.apache.iceberg.ContentScanTask; +import org.apache.iceberg.RewriteJobOrder; + +/** + * Container class representing a set of files to be rewritten by a {@link FileRewriteExecutor}. Review Comment: Let me apply this to the current situation, so we can check if I understand correctly: - We should keep `splitSize` in the group, as this drives the reading - We should move `expectedOutputFiles` to the plan, as it is driving the output Is this the change that you are suggesting? I see 3 possible separation of concerns: 1. Plan is the 'result' - everything below that is only organized based on the multiplicity of the data. So if some value applies to every group, then that value belongs to the 'global' plan variables. If a value is different for every group, then that value belongs to the group (current code) 2. Plan is the `write` config, group is the `read` config. If I understand correctly, this is what you are suggesting. IMHO this is a bit awkward, as currently the groups are part of the plan. Maybe we could have a `readConfig`, `writeConfig` map in the plan instead of adding extra attributes to the plan and to the groups. This comes with a cost of extra parsing for the configs, but allows us more flexibility (fewer classes) 3. The group should contain every information which is required for a single job. So the job (executor) only receives a single group and every other bit of information is global. The drawback is that some information is duplicated, but cleaner on the executor side. Your thoughts? -- 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: issues-unsubscr...@iceberg.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org --------------------------------------------------------------------- To unsubscribe, e-mail: issues-unsubscr...@iceberg.apache.org For additional commands, e-mail: issues-h...@iceberg.apache.org
