amogh-jahagirdar commented on code in PR #15006: URL: https://github.com/apache/iceberg/pull/15006#discussion_r2770148337
########## core/src/main/java/org/apache/iceberg/DVUtil.java: ########## @@ -0,0 +1,156 @@ +/* + * 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; + +import java.io.IOException; +import java.io.UncheckedIOException; +import java.util.List; +import java.util.Map; +import java.util.Objects; +import java.util.concurrent.ExecutorService; +import java.util.stream.Collectors; +import org.apache.iceberg.deletes.BaseDVFileWriter; +import org.apache.iceberg.deletes.DVFileWriter; +import org.apache.iceberg.deletes.Deletes; +import org.apache.iceberg.deletes.PositionDeleteIndex; +import org.apache.iceberg.io.DeleteWriteResult; +import org.apache.iceberg.io.OutputFileFactory; +import org.apache.iceberg.relocated.com.google.common.base.Preconditions; +import org.apache.iceberg.util.Tasks; +import org.slf4j.Logger; +import org.slf4j.LoggerFactory; + +class DVUtil { + private static final Logger LOG = LoggerFactory.getLogger(DVUtil.class); + + private DVUtil() {} + + /** + * Merges duplicate DVs for the same data file and writes the merged DV Puffin files. + * + * @param duplicateDVsByReferencedFile map of data file location to duplicate DVs (all entries + * must have size > 1) + * @return newly merged DVs + */ + static List<DeleteFile> mergeDVsAndWrite( + TableOperations ops, + Map<String, List<DeleteFile>> duplicateDVsByReferencedFile, + String tableName, + ExecutorService threadpool) { + Map<String, PositionDeleteIndex> mergedIndices = + duplicateDVsByReferencedFile.entrySet().stream() + .collect( + Collectors.toMap( + Map.Entry::getKey, + entry -> readDVsAndMerge(ops, entry.getValue(), threadpool))); Review Comment: I've updated this to read all the DVs that need to be merged in parallel rather than going group by group across the duplicates. As far as method signature goes, I thought about it some more and I still think that at the moment the right thing to pass is the map of referenced file to the list of DVs, so I'd prefer to leave it as is. The reason is before reading all the DVs we first need to figure out which DVs are duplicate, which neccessitates grouping by referenced file. Since MergingSnapshotProducer already tracks it by referenced file that algins pretty well. Another option is down the line we can also pass in just a List<DeleteFile> and defer the grouping to the utility but I don't think that's needed at the moment because we'd just have to regroup things (needlessly) again; we could add that when it's needed since it's package private. I could see this API being useful because we probably don't want to force a caller to have to compute the grouping. -- 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]
