yihua commented on code in PR #9883: URL: https://github.com/apache/hudi/pull/9883#discussion_r1373978234
########## hudi-client/hudi-spark-client/src/main/java/org/apache/hudi/merge/SparkRecordMergingUtils.java: ########## @@ -0,0 +1,192 @@ +/* + * 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.hudi.merge; + +import org.apache.hudi.AvroConversionUtils; +import org.apache.hudi.common.config.TypedProperties; +import org.apache.hudi.common.model.HoodieRecord; +import org.apache.hudi.common.model.HoodieRecordMerger; +import org.apache.hudi.common.model.HoodieSparkRecord; +import org.apache.hudi.common.util.collection.Pair; + +import org.apache.avro.Schema; +import org.apache.spark.sql.HoodieInternalRowUtils; +import org.apache.spark.sql.catalyst.InternalRow; +import org.apache.spark.sql.catalyst.expressions.GenericInternalRow; +import org.apache.spark.sql.types.StructField; +import org.apache.spark.sql.types.StructType; + +import java.util.ArrayList; +import java.util.Comparator; +import java.util.HashMap; +import java.util.List; +import java.util.Map; +import java.util.concurrent.ConcurrentHashMap; + +/** + * Util class to merge records that may contain partial updates. + * This can be plugged into any Spark {@link HoodieRecordMerger} implementation. + */ +public class SparkRecordMergingUtils { + private static final Map<Schema, Map<Integer, StructField>> FIELD_ID_TO_FIELD_MAPPING_CACHE = new ConcurrentHashMap<>(); + private static final Map<Schema, Map<String, Integer>> FIELD_NAME_TO_ID_MAPPING_CACHE = new ConcurrentHashMap<>(); + private static final Map<Pair<Schema, Schema>, + Pair<Map<Integer, StructField>, Pair<StructType, Schema>>> MERGED_SCHEMA_CACHE = new ConcurrentHashMap<>(); + + /** + * Merges records which can contain partial updates. + * + * @param older Older {@link HoodieSparkRecord}. + * @param oldSchema Old schema. + * @param newer Newer {@link HoodieSparkRecord}. + * @param newSchema New schema. + * @param props Configuration in {@link TypedProperties}. + * @return The merged record. + */ + public static Pair<HoodieRecord, Schema> mergeCompleteOrPartialRecords(HoodieSparkRecord older, + Schema oldSchema, + HoodieSparkRecord newer, + Schema newSchema, + TypedProperties props) { + Pair<Map<Integer, StructField>, Pair<StructType, Schema>> mappingSchemaPair = + getCachedMergedSchema(oldSchema, newSchema); + boolean isNewerPartial = isPartial(newSchema, mappingSchemaPair.getRight().getRight()); + if (isNewerPartial) { + InternalRow oldRow = older.getData(); + InternalRow newPartialRow = newer.getData(); + + Map<Integer, StructField> mergedIdToFieldMapping = mappingSchemaPair.getLeft(); + Map<String, Integer> newPartialNameToIdMapping = getCachedFieldNameToIdMapping(newSchema); + List<Object> values = new ArrayList<>(mergedIdToFieldMapping.size()); + for (int fieldId = 0; fieldId < mergedIdToFieldMapping.size(); fieldId++) { + StructField structField = mergedIdToFieldMapping.get(fieldId); + if (newPartialNameToIdMapping.containsKey(structField.name())) { + // pick from new + int ordInPartialUpdate = newPartialNameToIdMapping.get(structField.name()); + values.add(newPartialRow.get(ordInPartialUpdate, structField.dataType())); + } else { + // pick from old + values.add(oldRow.get(fieldId, structField.dataType())); + } + } + InternalRow mergedRow = new GenericInternalRow(values.toArray()); + + HoodieSparkRecord mergedSparkRecord = new HoodieSparkRecord( + mergedRow, mappingSchemaPair.getRight().getLeft()); + return Pair.of(mergedSparkRecord, mappingSchemaPair.getRight().getRight()); + } else { + return Pair.of(newer, newSchema); + } + } + + /** + * @param avroSchema Avro schema. + * @return The field ID to {@link StructField} instance mapping. + */ + public static Map<Integer, StructField> getCachedFieldIdToFieldMapping(Schema avroSchema) { + return FIELD_ID_TO_FIELD_MAPPING_CACHE.computeIfAbsent(avroSchema, schema -> { + StructType structType = HoodieInternalRowUtils.getCachedSchema(schema); + Map<Integer, StructField> schemaFieldIdMapping = new HashMap<>(); + int fieldId = 0; + + for (StructField field : structType.fields()) { + schemaFieldIdMapping.put(fieldId, field); + fieldId++; + } + + return schemaFieldIdMapping; + }); + } + + /** + * @param avroSchema Avro schema. + * @return The field name to ID mapping. + */ + public static Map<String, Integer> getCachedFieldNameToIdMapping(Schema avroSchema) { + return FIELD_NAME_TO_ID_MAPPING_CACHE.computeIfAbsent(avroSchema, schema -> { + StructType structType = HoodieInternalRowUtils.getCachedSchema(schema); + Map<String, Integer> schemaFieldIdMapping = new HashMap<>(); + int fieldId = 0; + + for (StructField field : structType.fields()) { + schemaFieldIdMapping.put(field.name(), fieldId); + fieldId++; + } + + return schemaFieldIdMapping; + }); + } + + /** + * Merges the two schemas so the merged schema contains all the fields from the two schemas, + * with the same ordering of fields, and the fields of the old schema comes first. + * + * @param oldSchema Old schema. + * @param newSchema New schema. + * @return The ID to {@link StructField} instance mapping of the merged schema, and the + * {@link StructType} and Avro schema of the merged schema. + */ + public static Pair<Map<Integer, StructField>, Pair<StructType, Schema>> getCachedMergedSchema(Schema oldSchema, + Schema newSchema) { + return MERGED_SCHEMA_CACHE.computeIfAbsent(Pair.of(oldSchema, newSchema), schemaPair -> { + Schema schema1 = schemaPair.getLeft(); + Schema schema2 = schemaPair.getRight(); + Map<Integer, StructField> idToFieldMapping1 = getCachedFieldIdToFieldMapping(schema1); + Map<Integer, StructField> idToFieldMapping2 = getCachedFieldIdToFieldMapping(schema2); + Map<String, Integer> nameToIdMapping1 = getCachedFieldNameToIdMapping(schema1); + Map<String, Integer> nameToIdMapping2 = getCachedFieldNameToIdMapping(schema2); + List<Integer> newFieldIdList = new ArrayList<>(); + for (String name : nameToIdMapping2.keySet()) { + if (!nameToIdMapping1.containsKey(name)) { + newFieldIdList.add(nameToIdMapping2.get(name)); + } + } + + if (newFieldIdList.isEmpty()) { + return Pair.of(idToFieldMapping1, Pair.of(HoodieInternalRowUtils.getCachedSchema(oldSchema), schema1)); + } else { + Map<Integer, StructField> mergedMapping = new HashMap<>(idToFieldMapping1); + int newFieldId = mergedMapping.size(); + newFieldIdList.sort(Comparator.naturalOrder()); + for (Integer fieldId : newFieldIdList) { + mergedMapping.put(newFieldId, idToFieldMapping2.get(fieldId)); + newFieldId++; + } + StructField[] fields = new StructField[mergedMapping.size()]; + for (int i = 0; i < fields.length; i++) { + fields[i] = mergedMapping.get(i); Review Comment: Yes. For no schema evolution, i.e., updating existing fields, there is no new fields to append in this part of logic. -- 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: commits-unsubscr...@hudi.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org
