How to use

addGivenOrTempPath

method

in

org.qcri.rheem.basic.channels.FileChannel$Instance

final String path = output.addGivenOrTempPath(this.targetPath, javaExecutor.getCompiler().getConfiguration());
final FileSystem fileSystem = FileSystems.getFileSystem(path).orElseThrow(
    () -> new IllegalStateException(String.format("No file system found for \"%s\".", this.targetPath))

final String path = output.addGivenOrTempPath(this.targetPath, javaExecutor.getCompiler().getConfiguration());

@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 FileChannel.Instance output = (FileChannel.Instance) outputs[0];
  final String targetPath = output.addGivenOrTempPath(this.targetPath, sparkExecutor.getConfiguration());
  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);
  LoggerFactory.getLogger(this.getClass()).info("Writing dataset to {}.", targetPath);
  return ExecutionOperator.modelEagerExecution(inputs, outputs, operatorContext);
}

@Override
public Tuple<Collection<ExecutionLineageNode>, Collection<ChannelInstance>> evaluate(
    ChannelInstance[] inputs,
    ChannelInstance[] outputs,
    SparkExecutor sparkExecutor,
    OptimizationContext.OperatorContext operatorContext) {
  assert inputs.length == this.getNumInputs();
  final FileChannel.Instance output = (FileChannel.Instance) outputs[0];
  final String targetPath = output.addGivenOrTempPath(this.targetPath, sparkExecutor.getConfiguration());
  final RddChannel.Instance input = (RddChannel.Instance) inputs[0];
  final JavaRDD<Object> rdd = input.provideRdd();
  final JavaRDD<String> serializedRdd = rdd
      .map(dataQuantum -> {
        // TODO: Once there are more tuple types, make this generic.
        @SuppressWarnings("unchecked")
        Tuple2<Object, Object> tuple2 = (Tuple2<Object, Object>) dataQuantum;
        return String.valueOf(tuple2.field0) + '\t' + String.valueOf(tuple2.field1);
      });
  this.name(serializedRdd);
  serializedRdd
      .coalesce(1) // TODO: Allow more than one TSV file?
      .saveAsTextFile(targetPath);
  return ExecutionOperator.modelEagerExecution(inputs, outputs, operatorContext);
}