ParquetProtobufBase.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.parquet.protobuf;
import com.google.protobuf.Message;
import com.google.protobuf.MessageOrBuilder;
import org.apache.flink.api.java.DataSet;
import org.apache.hadoop.fs.Path;
import org.apache.hadoop.mapreduce.Job;
import org.apache.hadoop.mapreduce.lib.input.FileInputFormat;
import org.apache.hadoop.mapreduce.lib.output.FileOutputFormat;
import org.apache.parquet.hadoop.ParquetFileWriter;
import org.apache.parquet.hadoop.ParquetOutputFormat;
import org.apache.parquet.hadoop.metadata.CompressionCodecName;
import org.apache.parquet.hadoop.util.ContextUtil;
import org.apache.parquet.proto.ProtoParquetInputFormat;
import org.apache.parquet.proto.ProtoReadSupport;
import org.apache.parquet.proto.ProtoWriteSupport;
import org.gradoop.flink.io.impl.parquet.common.HadoopValueInputFormat;
import org.gradoop.flink.io.impl.parquet.common.HadoopValueOutputFormat;
import org.gradoop.flink.io.impl.parquet.common.ParquetOutputFormatWithMode;
import org.gradoop.flink.io.impl.parquet.protobuf.kryo.ProtobufBuilderKryoSerializer;
import org.gradoop.flink.io.impl.parquet.protobuf.kryo.ProtobufMessageKryoSerializer;
import org.gradoop.flink.util.GradoopFlinkConfig;
import java.io.File;
import java.io.IOException;
import java.util.Objects;
/**
* Base class for parquet-protobuf data source and data sink.
*/
abstract class ParquetProtobufBase {
/**
* File ending for parquet-protobuf files.
*/
private static final String PROTOBUF_PARQUET_FILE_SUFFIX = ".proto.parquet";
/**
* parquet-protobuf file for vertices.
*/
private static final String VERTEX_FILE = "vertices" + PROTOBUF_PARQUET_FILE_SUFFIX;
/**
* parquet-protobuf file for graph heads.
*/
private static final String GRAPH_HEAD_FILE = "graphs" + PROTOBUF_PARQUET_FILE_SUFFIX;
/**
* parquet-protobuf file for edges.
*/
private static final String EDGE_FILE = "edges" + PROTOBUF_PARQUET_FILE_SUFFIX;
/**
* Root directory containing the CSV and metadata files.
*/
private final String basePath;
/**
* Gradoop Flink configuration
*/
private final GradoopFlinkConfig config;
/**
* Constructor.
*
* @param basePath directory to the parquet-protobuf files
* @param config Gradoop Flink configuration
*/
protected ParquetProtobufBase(String basePath, GradoopFlinkConfig config) {
Objects.requireNonNull(basePath);
Objects.requireNonNull(config);
this.basePath = basePath.endsWith(File.separator) ? basePath : basePath + File.separator;
this.config = config;
// see: https://github.com/apache/flink/pull/7865
config.getExecutionEnvironment()
.addDefaultKryoSerializer(Message.Builder.class, ProtobufBuilderKryoSerializer.class);
config.getExecutionEnvironment()
.addDefaultKryoSerializer(Message.class, ProtobufMessageKryoSerializer.class);
}
/**
* Returns the path to the graph head file.
*
* @return graph head file path
*/
protected String getGraphHeadPath() {
return basePath + GRAPH_HEAD_FILE;
}
/**
* Returns the path to the vertex file.
*
* @return vertex file path
*/
protected String getVertexPath() {
return basePath + VERTEX_FILE;
}
/**
* Returns the path to the edge file.
*
* @return edge file path
*/
protected String getEdgePath() {
return basePath + EDGE_FILE;
}
/**
* Get the gradoop configuration.
*
* @return the gradoop configuration instance
*/
protected GradoopFlinkConfig getConfig() {
return config;
}
/**
* Writes a protobuf {@link DataSet} as parquet-protobuf file(s) to the specified location.
*
* @param data the protobuf {@link DataSet}
* @param type the protobuf object class
* @param outputPath the path of the file
* @param overwrite the behavior regarding existing files
* @param <T> protobuf object type
* @throws IOException if an I/O error occurs
*/
public <T extends MessageOrBuilder> void write(DataSet<T> data, Class<? extends Message> type,
String outputPath, boolean overwrite) throws IOException {
Job job = Job.getInstance();
FileOutputFormat.setOutputPath(job, new Path(outputPath));
ParquetOutputFormat.setCompression(job, CompressionCodecName.SNAPPY);
ParquetOutputFormat.setEnableDictionary(job, false);
ParquetOutputFormat.setWriteSupportClass(job, ProtoWriteSupport.class);
ParquetOutputFormatWithMode.setFileCreationMode(job,
overwrite ? ParquetFileWriter.Mode.OVERWRITE : ParquetFileWriter.Mode.CREATE);
ProtoWriteSupport.setSchema(ContextUtil.getConfiguration(job), type);
ProtoWriteSupport.setWriteSpecsCompliant(ContextUtil.getConfiguration(job), true);
data.output(new HadoopValueOutputFormat<T>(new ParquetOutputFormatWithMode<>(), job));
}
/**
* Creates a {@link DataSet} that represents the protobuf objects produced by reading the given
* parquet-protobuf file.
*
* @param type the protobuf object class
* @param inputPath the path of the file
* @return A {@link DataSet} that represents the data read from the given file.
* @param <T> protobuf object type
* @throws IOException if an I/O error occurs
*/
public <T extends Message.Builder> DataSet<T> read(Class<T> type, String inputPath) throws
IOException {
Job job = Job.getInstance();
FileInputFormat.addInputPath(job, new Path(inputPath));
ProtoReadSupport.setProtobufClass(ContextUtil.getConfiguration(job),
type.getDeclaringClass().asSubclass(Message.class).getName());
return config.getExecutionEnvironment()
.createInput(new HadoopValueInputFormat<T>(new ProtoParquetInputFormat<>(), type, job));
}
}