Java Code Examples for org.apache.avro.file.DataFileWriter#setMeta()

@SuppressWarnings("unchecked")
private static <D> DataFileWriter<D> newAvroWriter(
    Schema schema, OutputFile file, Function<Schema, DatumWriter<?>> createWriterFunc,
    CodecFactory codec, Map<String, String> metadata) throws IOException {
  DataFileWriter<D> writer = new DataFileWriter<>(
      (DatumWriter<D>) createWriterFunc.apply(schema));

  writer.setCodec(codec);

  for (Map.Entry<String, String> entry : metadata.entrySet()) {
    writer.setMeta(entry.getKey(), entry.getValue());
  }

  // TODO: support overwrite
  return writer.create(schema, file.create());
}
 

@Override
public void open(WritableByteChannel channel) throws IOException {
  this.schema = new Schema.Parser().parse(getJsonSchema());
  DataFileWriter<?> writer;
  if (getRecordFormatter() == null) {
    writer = reflectWriter = new DataFileWriter<>(new ReflectDatumWriter<>(schema));
  } else {
    writer = genericWriter = new DataFileWriter<>(new GenericDatumWriter<>(schema));
  }
  writer.setCodec(getCodec().getCodec());
  for (Map.Entry<String, Object> entry : getMetadata().entrySet()) {
    Object v = entry.getValue();
    if (v instanceof String) {
      writer.setMeta(entry.getKey(), (String) v);
    } else if (v instanceof Long) {
      writer.setMeta(entry.getKey(), (Long) v);
    } else if (v instanceof byte[]) {
      writer.setMeta(entry.getKey(), (byte[]) v);
    } else {
      throw new IllegalStateException(
          "Metadata value type must be one of String, Long, or byte[]. Found "
              + v.getClass().getSimpleName());
    }
  }
  writer.create(schema, Channels.newOutputStream(channel));
}
 

static <K> void configureDataFileWriter(DataFileWriter<K> writer,
    JobConf job) throws UnsupportedEncodingException {

    if (FileOutputFormat.getCompressOutput(job)) {
        int level = job.getInt(org.apache.avro.mapred.AvroOutputFormat.DEFLATE_LEVEL_KEY,
                org.apache.avro.mapred.AvroOutputFormat.DEFAULT_DEFLATE_LEVEL);
        String codecName = job.get(AvroJob.OUTPUT_CODEC, DEFLATE_CODEC);
        CodecFactory factory = codecName.equals(DEFLATE_CODEC) ?
            CodecFactory.deflateCodec(level) : CodecFactory.fromString(codecName);
        writer.setCodec(factory);
    }

    writer.setSyncInterval(job.getInt(org.apache.avro.mapred.AvroOutputFormat.SYNC_INTERVAL_KEY,
            DEFAULT_SYNC_INTERVAL));

    // copy metadata from job
    for (Map.Entry<String,String> e : job) {
        if (e.getKey().startsWith(AvroJob.TEXT_PREFIX))
            writer.setMeta(e.getKey().substring(AvroJob.TEXT_PREFIX.length()),e.getValue());
        if (e.getKey().startsWith(AvroJob.BINARY_PREFIX))
            writer.setMeta(e.getKey().substring(AvroJob.BINARY_PREFIX.length()),
                   URLDecoder.decode(e.getValue(), "ISO-8859-1")
                   .getBytes("ISO-8859-1"));
    }
}
 

private byte[] initAvroWriter(ProcessSession session, String codec, DataFileStream<GenericRecord> reader,
                                     DataFileWriter<GenericRecord> writer, AtomicReference<FlowFile> flowFileRef) {

    writer.setCodec(CodecFactory.fromString(codec));
    // Transfer metadata (this is a subset of the incoming file)
    for (String metaKey : reader.getMetaKeys()) {
        if (!RESERVED_METADATA.contains(metaKey)) {
            writer.setMeta(metaKey, reader.getMeta(metaKey));
        }
    }

    final ByteArrayOutputStream avroHeader = new ByteArrayOutputStream();
    flowFileRef.set(session.append(flowFileRef.get(), (out) -> {
        // Create writer so that records can be appended later.
        writer.create(reader.getSchema(), avroHeader);
        writer.close();

        final byte[] header = avroHeader.toByteArray();
        out.write(header);
    }));

    // Capture the Avro header byte array that is just written to the FlowFile.
    // This is needed when Avro records are appended to the same FlowFile.
    return avroHeader.toByteArray();
}
 

@SuppressWarnings("unchecked")
private static <D> DataFileWriter<D> newAvroWriter(
    Schema schema, PositionOutputStream stream, Function<Schema, DatumWriter<?>> createWriterFunc,
    CodecFactory codec, Map<String, String> metadata) throws IOException {
  DataFileWriter<D> writer = new DataFileWriter<>(
      (DatumWriter<D>) createWriterFunc.apply(schema));

  writer.setCodec(codec);

  for (Map.Entry<String, String> entry : metadata.entrySet()) {
    writer.setMeta(entry.getKey(), entry.getValue());
  }

  return writer.create(schema, stream);
}
 

static <T> void configureDataFileWriter(DataFileWriter<T> writer,
  TaskAttemptContext context) throws UnsupportedEncodingException {
  if (FileOutputFormat.getCompressOutput(context)) {
    int level = context.getConfiguration()
      .getInt(DEFLATE_LEVEL_KEY, DEFAULT_DEFLATE_LEVEL);
    String codecName = context.getConfiguration()
      .get(org.apache.avro.mapred.AvroJob.OUTPUT_CODEC, DEFLATE_CODEC);
    CodecFactory factory =
      codecName.equals(DEFLATE_CODEC) ? CodecFactory.deflateCodec(level)
        : CodecFactory.fromString(codecName);
    writer.setCodec(factory);
  }

  writer.setSyncInterval(context.getConfiguration()
    .getInt(SYNC_INTERVAL_KEY, DEFAULT_SYNC_INTERVAL));

  // copy metadata from job
  for (Map.Entry<String, String> e : context.getConfiguration()) {
    if (e.getKey().startsWith(org.apache.avro.mapred.AvroJob.TEXT_PREFIX)) {
      writer.setMeta(e.getKey()
        .substring(org.apache.avro.mapred.AvroJob.TEXT_PREFIX.length()),
        e.getValue());
    }
    if (e.getKey().startsWith(org.apache.avro.mapred.AvroJob.BINARY_PREFIX)) {
      writer.setMeta(e.getKey()
        .substring(org.apache.avro.mapred.AvroJob.BINARY_PREFIX.length()),
        URLDecoder.decode(e.getValue(), "ISO-8859-1").getBytes("ISO-8859-1"));
    }
  }
}
 

static void configureDataFileWriter(DataFileWriter<GenericData.Record> writer,
    JobConf job) throws UnsupportedEncodingException {
  if (FileOutputFormat.getCompressOutput(job)) {
    int level = job.getInt(DEFLATE_LEVEL_KEY,
        DEFAULT_DEFLATE_LEVEL);
    String codecName = job.get(AvroJob.OUTPUT_CODEC, DEFLATE_CODEC);
    CodecFactory factory = codecName.equals(DEFLATE_CODEC)
      ? CodecFactory.deflateCodec(level)
      : CodecFactory.fromString(codecName);
    writer.setCodec(factory);
  }

  // Do max as core-default.xml has io.file.buffer.size as 4K
  writer.setSyncInterval(job.getInt(SYNC_INTERVAL_KEY, Math.max(
          job.getInt("io.file.buffer.size", DEFAULT_SYNC_INTERVAL), DEFAULT_SYNC_INTERVAL)));

  // copy metadata from job
  for (Map.Entry<String,String> e : job) {
    if (e.getKey().startsWith(AvroJob.TEXT_PREFIX))
      writer.setMeta(e.getKey().substring(AvroJob.TEXT_PREFIX.length()),
                     e.getValue());
    if (e.getKey().startsWith(AvroJob.BINARY_PREFIX))
      writer.setMeta(e.getKey().substring(AvroJob.BINARY_PREFIX.length()),
                     URLDecoder.decode(e.getValue(), "ISO-8859-1")
                     .getBytes("ISO-8859-1"));
  }
}
 

private void createOutput(DateRange dateRange) throws IOException
{
  DataFileWriter<GenericRecord> dataWriter;
  OutputStream outputStream;
  
  Path path = new Path(_outputPath,PathUtils.datedPathFormat.format(dateRange.getEndDate()));
  
  Schema ouputSchema = Schemas.createRecordSchema(PartitionCollapsingTests.class, "Output",
                                            new Field("id", Schema.create(Type.LONG), "ID", null));
  
  outputStream = getFileSystem().create(new Path(path, "part-00000.avro"));
  
  GenericDatumWriter<GenericRecord> writer = new GenericDatumWriter<GenericRecord>();
  dataWriter = new DataFileWriter<GenericRecord>(writer);      
  
  dataWriter.setMeta(AvroDateRangeMetadata.METADATA_DATE_START,
                     Long.toString(dateRange.getBeginDate().getTime()));
  
  dataWriter.setMeta(AvroDateRangeMetadata.METADATA_DATE_END,
                     Long.toString(dateRange.getEndDate().getTime()));
  
  dataWriter.create(ouputSchema, outputStream);
      
  // empty file
  
  dataWriter.close();
  outputStream.close();
  dataWriter = null;
  outputStream = null; 
}