org.apache.crunch.MapFn Java Examples

The following examples show how to use org.apache.crunch.MapFn. 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: CassandraParams.java    From hdfs2cass with Apache License 2.0 5 votes vote down vote up 
private static MapFn<CQLRecord, ByteBuffer> makeKeyFn(final int[] partitionKeyIndexes) {
  return new MapFn<CQLRecord, ByteBuffer>() {
    @Override
    public ByteBuffer map(final CQLRecord record) {
      return CassandraRecordUtils.getPartitionKey(record.getValues(), partitionKeyIndexes);
    }
  };
}
Example #2
Source File: CassandraParams.java    From hdfs2cass with Apache License 2.0 4 votes vote down vote up 
/**
 * @return a map function to extract the partition key from a record
 */
public MapFn<CQLRecord, ByteBuffer> getKeyFn() {
  return makeKeyFn(clusterInfo.getPartitionKeyIndexes());
}
Example #3
Source File: TransformTask.java    From kite with Apache License 2.0 4 votes vote down vote up 
public PipelineResult run() throws IOException {
  boolean isLocal = (isLocal(from.getDataset()) || isLocal(to.getDataset()));
  if (isLocal) {
    // copy to avoid making changes to the caller's configuration
    Configuration conf = new Configuration(getConf());
    conf.set("mapreduce.framework.name", "local");
    setConf(conf);
  }

  if (isHive(from) || isHive(to)) {
    setConf(addHiveDelegationToken(getConf()));

    // add jars needed for metastore interaction to the classpath
    if (!isLocal) {
      Class<?> fb303Class, thriftClass;
      try {
        // attempt to use libfb303 and libthrift 0.9.2 when async was added
        fb303Class = Class.forName(
            "com.facebook.fb303.FacebookService.AsyncProcessor");
        thriftClass = Class.forName(
            "org.apache.thrift.TBaseAsyncProcessor");
      } catch (ClassNotFoundException e) {
        try {
          // fallback to 0.9.0 or earlier
          fb303Class = Class.forName(
              "com.facebook.fb303.FacebookBase");
          thriftClass = Class.forName(
              "org.apache.thrift.TBase");
        } catch (ClassNotFoundException real) {
          throw new DatasetOperationException(
              "Cannot find thrift dependencies", real);
        }
      }

      TaskUtil.configure(getConf())
          .addJarForClass(Encoder.class) // commons-codec
          .addJarForClass(Log.class) // commons-logging
          .addJarForClass(CompressorInputStream.class) // commons-compress
          .addJarForClass(ApiAdapter.class) // datanucleus-core
          .addJarForClass(JDOAdapter.class) // datanucleus-api-jdo
          .addJarForClass(SQLQuery.class) // datanucleus-rdbms
          .addJarForClass(JDOHelper.class) // jdo-api
          .addJarForClass(Transaction.class) // jta
          .addJarForClass(fb303Class) // libfb303
          .addJarForClass(thriftClass) // libthrift
          .addJarForClass(HiveMetaStore.class) // hive-metastore
          .addJarForClass(HiveConf.class); // hive-exec
    }
  }

  PType<T> toPType = ptype(to);
  MapFn<T, T> validate = new CheckEntityClass<T>(to.getType());

  Pipeline pipeline = new MRPipeline(getClass(), getConf());

  PCollection<T> collection = pipeline.read(CrunchDatasets.asSource(from))
      .parallelDo(transform, toPType).parallelDo(validate, toPType);

  if (compact) {
    // the transform must be run before partitioning
    collection = CrunchDatasets.partition(collection, to, numWriters, numPartitionWriters);
  }

  pipeline.write(collection, CrunchDatasets.asTarget(to), mode);

  PipelineResult result = pipeline.done();

  StageResult sr = Iterables.getFirst(result.getStageResults(), null);
  if (sr != null && MAP_INPUT_RECORDS != null) {
    this.count = sr.getCounterValue(MAP_INPUT_RECORDS);
  }

  return result;
}