Java Code Examples for org.apache.beam.sdk.io.TextIO#Write

The following examples show how to use org.apache.beam.sdk.io.TextIO#Write . You can vote up the ones you like or vote down the ones you don't like, and go to the original project or source file by following the links above each example. You may check out the related API usage on the sidebar.
Example 1
Source File: InvoicingPipeline.java    From nomulus with Apache License 2.0 6 votes vote down vote up 
/** Returns an IO transform that writes the overall invoice to a single CSV file. */
private TextIO.Write writeInvoice(ValueProvider<String> yearMonthProvider) {
  return TextIO.write()
      .to(
          NestedValueProvider.of(
              yearMonthProvider,
              yearMonth ->
                  String.format(
                      "%s/%s/%s/%s-%s",
                      billingBucketUrl,
                      BillingModule.INVOICES_DIRECTORY,
                      yearMonth,
                      invoiceFilePrefix,
                      yearMonth)))
      .withHeader(InvoiceGroupingKey.invoiceHeader())
      .withoutSharding()
      .withSuffix(".csv");
}
Example 2
Source File: DynamicOneFilePerWindow.java    From dlp-dataflow-deidentification with Apache License 2.0 5 votes vote down vote up 
@Override
public PDone expand(PCollection<KV<String, String>> input) {

  PCollection<String> contents =
      input.apply(
          ParDo.of(
              new DoFn<KV<String, String>, String>() {
                @ProcessElement
                public void processElement(ProcessContext c) {
                  filenamePrefix = String.format("%s%s", filenamePrefix, c.element().getKey());
                  LOG.info("File Prefix {}", filenamePrefix);

                  c.output(c.element().getValue());
                }
              }));

  ResourceId resource = FileBasedSink.convertToFileResourceIfPossible(filenamePrefix);
  TextIO.Write write =
      TextIO.write()
          .to(new PerWindowFiles(resource))
          .withTempDirectory(resource.getCurrentDirectory())
          .withWindowedWrites();

  if (numShards != null) {
    write = write.withNumShards(numShards);
  }

  return contents.apply(write);
}
Example 3
Source File: WriteOneFilePerWindow.java    From dlp-dataflow-deidentification with Apache License 2.0 5 votes vote down vote up 
@Override
public PDone expand(PCollection<String> input) {

  ResourceId resource = FileBasedSink.convertToFileResourceIfPossible(filenamePrefix);
  TextIO.Write write =
      TextIO.write()
          .to(new PerWindowFiles(resource))
          .withTempDirectory(resource.getCurrentDirectory())
          .withWindowedWrites();

  if (numShards != null) {
    write = write.withNumShards(numShards);
  }

  return input.apply(write);
}
Example 4
Source File: WriteOneFilePerWindow.java    From incubator-nemo with Apache License 2.0 5 votes vote down vote up 
@Override
public PDone expand(final PCollection<String> input) {
  final ResourceId resource = FileBasedSink.convertToFileResourceIfPossible(filenamePrefix);
  TextIO.Write write =
    TextIO.write()
      .to(new PerWindowFiles(resource))
      .withTempDirectory(resource.getCurrentDirectory())
      .withWindowedWrites();
  if (numShards != null) {
    write = write.withNumShards(numShards);
  }
  return input.apply(write);
}
Example 5
Source File: WriteOneFilePerWindow.java    From deployment-examples with MIT License 5 votes vote down vote up 
@Override
public PDone expand(PCollection<String> input) {
  ResourceId resource = FileBasedSink.convertToFileResourceIfPossible(filenamePrefix);
  TextIO.Write write =
      TextIO.write()
          .to(new PerWindowFiles(resource))
          .withTempDirectory(resource.getCurrentDirectory())
          .withWindowedWrites();
  if (numShards != null) {
    write = write.withNumShards(numShards);
  }
  return input.apply(write);
}
Example 6
Source File: WriteOneFilePerWindow.java    From beam with Apache License 2.0 5 votes vote down vote up 
@Override
public PDone expand(PCollection<String> input) {
  ResourceId resource = FileBasedSink.convertToFileResourceIfPossible(filenamePrefix);
  TextIO.Write write =
      TextIO.write()
          .to(new PerWindowFiles(resource))
          .withTempDirectory(resource.getCurrentDirectory())
          .withWindowedWrites();
  if (numShards != null) {
    write = write.withNumShards(numShards);
  }
  return input.apply(write);
}
Example 7
Source File: BeamOutputTransform.java    From hop with Apache License 2.0 4 votes vote down vote up 
@Override public PDone expand( PCollection<HopRow> input ) {

    try {
      // Only initialize once on this node/vm
      //
      BeamHop.init(transformPluginClasses, xpPluginClasses);

      // Inflate the metadata on the node where this is running...
      //
      IRowMeta rowMeta = JsonRowMeta.fromJson( rowMetaJson );

      // This is the end of a computing chain, we write out the results
      // We write a bunch of Strings, one per line basically
      //
      PCollection<String> stringCollection = input.apply( transformName, ParDo.of( new HopToStringFn( transformName, outputLocation, separator, enclosure, rowMetaJson, transformPluginClasses, xpPluginClasses ) ) );

      // We need to transform these lines into a file and then we're PDone
      //
      TextIO.Write write = TextIO.write();
      if ( StringUtils.isNotEmpty(outputLocation)) {
        String outputPrefix = outputLocation;
        if (!outputPrefix.endsWith( File.separator)) {
          outputPrefix+=File.separator;
        }
        if (StringUtils.isNotEmpty( filePrefix )) {
          outputPrefix+=filePrefix;
        }
        write = write.to( outputPrefix );
      }
      if (StringUtils.isNotEmpty( fileSuffix )) {
        write = write.withSuffix( fileSuffix );
      }

      // For streaming data sources...
      //
      if (windowed) {
        write = write.withWindowedWrites().withNumShards( 4 ); // TODO config
      }

      stringCollection.apply(write);

      // Get it over with
      //
      return PDone.in(input.getPipeline());

    } catch ( Exception e ) {
      numErrors.inc();
      LOG.error( "Error in beam output transform", e );
      throw new RuntimeException( "Error in beam output transform", e );
    }
  }
Example 8
Source File: BeamOutputTransform.java    From kettle-beam with Apache License 2.0 4 votes vote down vote up 
@Override public PDone expand( PCollection<KettleRow> input ) {

    try {
      // Only initialize once on this node/vm
      //
      BeamKettle.init(stepPluginClasses, xpPluginClasses);

      // Inflate the metadata on the node where this is running...
      //
      RowMetaInterface rowMeta = JsonRowMeta.fromJson( rowMetaJson );

      // This is the end of a computing chain, we write out the results
      // We write a bunch of Strings, one per line basically
      //
      PCollection<String> stringCollection = input.apply( stepname, ParDo.of( new KettleToStringFn( stepname, outputLocation, separator, enclosure, rowMetaJson, stepPluginClasses, xpPluginClasses ) ) );

      // We need to transform these lines into a file and then we're PDone
      //
      TextIO.Write write = TextIO.write();
      if ( StringUtils.isNotEmpty(outputLocation)) {
        String outputPrefix = outputLocation;
        if (!outputPrefix.endsWith( File.separator)) {
          outputPrefix+=File.separator;
        }
        if (StringUtils.isNotEmpty( filePrefix )) {
          outputPrefix+=filePrefix;
        }
        write = write.to( outputPrefix );
      }
      if (StringUtils.isNotEmpty( fileSuffix )) {
        write = write.withSuffix( fileSuffix );
      }

      // For streaming data sources...
      //
      if (windowed) {
        write = write.withWindowedWrites().withNumShards( 4 ); // TODO config
      }

      stringCollection.apply(write);

      // Get it over with
      //
      return PDone.in(input.getPipeline());

    } catch ( Exception e ) {
      numErrors.inc();
      LOG.error( "Error in beam output transform", e );
      throw new RuntimeException( "Error in beam output transform", e );
    }
  }
Example 9
Source File: TextTableProvider.java    From beam with Apache License 2.0 4 votes vote down vote up 
private TextIO.Write writeJsonToDlf() {
  return TextIO.write().withDelimiter(new char[] {}).to(deadLetterFile());
}