blob: 42b1ebbb54ce50fb492714a799b2563d08a6aeb6 [file] [log] [blame]
/*
* 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.wayang.spark.operators;
import java.util.Collection;
import java.util.Collections;
import java.util.List;
import org.apache.logging.log4j.LogManager;
import org.apache.wayang.basic.channels.FileChannel;
import org.apache.wayang.basic.operators.ObjectFileSink;
import org.apache.wayang.core.optimizer.OptimizationContext;
import org.apache.wayang.core.plan.wayangplan.ExecutionOperator;
import org.apache.wayang.core.plan.wayangplan.Operator;
import org.apache.wayang.core.platform.ChannelDescriptor;
import org.apache.wayang.core.platform.ChannelInstance;
import org.apache.wayang.core.platform.lineage.ExecutionLineageNode;
import org.apache.wayang.core.types.DataSetType;
import org.apache.wayang.core.util.Tuple;
import org.apache.wayang.spark.channels.RddChannel;
import org.apache.wayang.spark.execution.SparkExecutor;
import org.apache.wayang.spark.platform.SparkPlatform;
/**
* {@link Operator} for the {@link SparkPlatform} that creates a sequence file.
*
* @see SparkObjectFileSource
*/
public class SparkObjectFileSink<T> extends ObjectFileSink<T> implements SparkExecutionOperator {
public SparkObjectFileSink(ObjectFileSink<T> that) {
super(that);
}
public SparkObjectFileSink(DataSetType<T> type) {
this(null, type);
}
public SparkObjectFileSink(String targetPath, DataSetType<T> type) {
super(targetPath, type);
}
@Override
public Tuple<Collection<ExecutionLineageNode>, Collection<ChannelInstance>> evaluate(
ChannelInstance[] inputs,
ChannelInstance[] outputs,
SparkExecutor sparkExecutor,
OptimizationContext.OperatorContext operatorContext) {
assert inputs.length == this.getNumInputs();
assert outputs.length <= 1;
final String targetPath;
if(outputs.length > 0) {
final FileChannel.Instance output = (FileChannel.Instance) outputs[0];
targetPath = output.addGivenOrTempPath(this.textFileUrl, sparkExecutor.getConfiguration());
}else{
targetPath = this.textFileUrl;
}
RddChannel.Instance input = (RddChannel.Instance) inputs[0];
input.provideRdd()
.coalesce(1) // TODO: Remove. This only hotfixes the issue that JavaObjectFileSource reads only a single file.
.saveAsObjectFile(targetPath);
LogManager.getLogger(this.getClass()).info("Writing dataset to {}.", targetPath);
return ExecutionOperator.modelEagerExecution(inputs, outputs, operatorContext);
}
@Override
protected ExecutionOperator createCopy() {
return new SparkObjectFileSink<>(this.textFileUrl, this.getType());
}
@Override
public String getLoadProfileEstimatorConfigurationKey() {
return "wayang.spark.objectfilesink.load";
}
@Override
public List<ChannelDescriptor> getSupportedInputChannels(int index) {
return Collections.singletonList(RddChannel.UNCACHED_DESCRIPTOR);
}
@Override
public List<ChannelDescriptor> getSupportedOutputChannels(int index) {
return Collections.singletonList(FileChannel.HDFS_OBJECT_FILE_DESCRIPTOR);
}
@Override
public boolean containsAction() {
return true;
}
}