IndexedCSVDataSource.java

/*
 * Copyright © 2014 - 2021 Leipzig University (Database Research Group)
 *
 * Licensed 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.gradoop.flink.io.impl.csv.indexed;

import org.apache.flink.api.java.DataSet;
import org.apache.flink.api.java.ExecutionEnvironment;
import org.apache.flink.api.java.tuple.Tuple2;
import org.apache.flink.api.java.tuple.Tuple3;
import org.apache.hadoop.conf.Configuration;
import org.gradoop.common.model.impl.pojo.EPGMEdge;
import org.gradoop.common.model.impl.pojo.EPGMGraphHead;
import org.gradoop.common.model.impl.pojo.EPGMVertex;
import org.gradoop.flink.io.api.DataSource;
import org.gradoop.flink.io.impl.csv.CSVBase;
import org.gradoop.flink.io.impl.csv.functions.CSVLineToEdge;
import org.gradoop.flink.io.impl.csv.functions.CSVLineToGraphHead;
import org.gradoop.flink.io.impl.csv.functions.CSVLineToVertex;
import org.gradoop.flink.io.impl.csv.metadata.CSVMetaData;
import org.gradoop.flink.io.impl.csv.metadata.CSVMetaDataSource;
import org.gradoop.flink.model.impl.epgm.GraphCollection;
import org.gradoop.flink.model.impl.epgm.GraphCollectionFactory;
import org.gradoop.flink.model.impl.epgm.LogicalGraph;
import org.gradoop.flink.util.GradoopFlinkConfig;

import java.io.IOException;
import java.util.Map;
import java.util.Objects;
import java.util.stream.Collectors;

/**
 * A graph data source for CSV files indexed by label.
 * <p>
 * The datasource expects files separated by label, e.g. in the following directory structure:
 * <p>
 * csvRoot
 * |- Person.csv     # contains all vertices with label 'Person'
 * |- University.csv # contains all vertices with label 'University'
 * |- knows.csv      # contains all edges with label 'knows'
 * |- studyAt.csv    # contains all edges with label 'studyAt'
 * |- metadata.csv   # Meta data for all data contained in the graph
 */
public class IndexedCSVDataSource extends CSVBase implements DataSource {
  /**
   * HDFS Configuration.
   */
  private final Configuration hdfsConfig;

  /**
   * Creates a new data source. The constructor creates a default HDFS configuration.
   *
   * @param csvPath root path of csv files
   * @param config  gradoop configuration
   */
  public IndexedCSVDataSource(String csvPath, GradoopFlinkConfig config) {
    this(csvPath, config, new Configuration());
  }

  /**
   * Creates a new data source.
   *
   * @param csvPath  root path of csv files
   * @param conf     gradoop configuration
   * @param hdfsConf HDFS configuration
   */
  public IndexedCSVDataSource(String csvPath, GradoopFlinkConfig conf, Configuration hdfsConf) {
    super(csvPath, conf);
    this.hdfsConfig = Objects.requireNonNull(hdfsConf);
  }

  /**
   * Will use a single graph head of the collection as final graph head for the graph.
   * Issue #1217 (https://github.com/dbs-leipzig/gradoop/issues/1217) will optimize further.
   *
   * {@inheritDoc}
   */
  @Override
  public LogicalGraph getLogicalGraph() throws IOException {
    GraphCollection collection = getGraphCollection();
    return collection.getGraphFactory()
      .fromDataSets(collection.getGraphHeads().first(1), collection.getVertices(), collection.getEdges());
  }

  @Override
  public GraphCollection getGraphCollection() throws IOException {
    CSVMetaDataSource source = new CSVMetaDataSource();
    CSVMetaData metaData = source.readLocal(getMetaDataPath(), hdfsConfig);
    DataSet<Tuple3<String, String, String>> metaDataBroadcast =
      source.readDistributed(getMetaDataPath(), getConfig());

    ExecutionEnvironment env = getConfig().getExecutionEnvironment();
    GraphCollectionFactory factory = getConfig().getGraphCollectionFactory();

    Map<String, DataSet<EPGMGraphHead>> graphHeads = metaData.getGraphLabels().stream()
      .map(label -> Tuple2.of(label, env.readTextFile(getGraphHeadCSVPath(label))
        .map(new CSVLineToGraphHead(factory.getGraphHeadFactory()))
        .withBroadcastSet(metaDataBroadcast, BC_METADATA)
        .filter(graphHead -> graphHead.getLabel().equals(label))))
      .collect(Collectors.toMap(t -> t.f0, t -> t.f1));

    Map<String, DataSet<EPGMVertex>> vertices = metaData.getVertexLabels().stream()
      .map(label -> Tuple2.of(label, env.readTextFile(getVertexCSVPath(label))
        .map(new CSVLineToVertex(factory.getVertexFactory()))
        .withBroadcastSet(metaDataBroadcast, BC_METADATA)
        .filter(vertex -> vertex.getLabel().equals(label))))
      .collect(Collectors.toMap(t -> t.f0, t -> t.f1));

    Map<String, DataSet<EPGMEdge>> edges = metaData.getEdgeLabels().stream()
      .map(label -> Tuple2.of(label, env.readTextFile(getEdgeCSVPath(label))
        .map(new CSVLineToEdge(factory.getEdgeFactory()))
        .withBroadcastSet(metaDataBroadcast, BC_METADATA)
        .filter(edge -> edge.getLabel().equals(label))))
      .collect(Collectors.toMap(t -> t.f0, t -> t.f1));

    return factory.fromIndexedDataSets(graphHeads, vertices, edges);
  }
}