Examples of cascading.tuple.Tuple

cascading.tuple.Tuple
A Tuple represents a set of values. Consider a Tuple the same as a database record where every value is a column in that table.
A "tuple stream" is a set of Tuple instances passed consecutively through a Pipe assembly.
Tuples work in tandem with {@link Fields} and the {@link TupleEntry} classes. A TupleEntry holds an instance ofFields and a Tuple. It allows a tuple to be accessed by its field names, and will help maintain consistent types if any are given on the Fields instance. That is, if a field is specified at an Integer, calling {@link #set(int,Object)}with a String will force the String to be coerced into a Integer instance.
For managing custom types, see the {@link CoercibleType} interface which extends {@link Type}.
Tuple instances created by user code, by default, are mutable (or modifiable). Tuple instances created by the system are immutable (or unmodifiable, tested by calling {@link #isUnmodifiable()}).
For example tuples returned by {@link cascading.operation.FunctionCall#getArguments()}, will always be unmodifiable. Thus they must be copied if they will be changed by user code or cached in the local context. See the Tuple copy constructor, or {@code *Copy()} methodson {@link TupleEntry}.
Because a Tuple can hold any Object type, it is suitable for storing custom types. But all custom types must have a serialization support per the underlying framework.
For Hadoop, a {@link org.apache.hadoop.io.serializer.Serialization} implementationmust be registered with Hadoop. For further performance improvements, see the {@link cascading.tuple.hadoop.SerializationToken} Java annotation. @see org.apache.hadoop.io.serializer.Serialization @see cascading.tuple.hadoop.SerializationToken

  @Override
  public void sink(FlowProcess<JobConf> flowProcess, SinkCall<Object[], OutputCollector> sinkCall)
      throws IOException {
    TupleEntry tupleEntry = sinkCall.getOutgoingEntry();
    OutputCollector outputCollector = sinkCall.getOutput();
    Tuple key = tupleEntry.selectTuple(keyField);
    ImmutableBytesWritable keyBytes = (ImmutableBytesWritable) key.getObject(0);


    if (useSalt) {
      keyBytes = HBaseSalter.addSaltPrefix(keyBytes);
    }


    Put put;
    if (this.timeStamp == 0L) {
      put = new Put(keyBytes.get());
    } else {
      put = new Put(keyBytes.get(), this.timeStamp);
    }
    
    for (int i = 0; i < valueFields.length; i++) {
      Fields fieldSelector = valueFields[i];
      TupleEntry values = tupleEntry.selectEntry(fieldSelector);


      for (int j = 0; j < values.getFields().size(); j++) {
        Fields fields = values.getFields();
        Tuple tuple = values.getTuple();


        ImmutableBytesWritable valueBytes = (ImmutableBytesWritable) tuple.getObject(j);
        if (valueBytes != null)
            put.add(Bytes.toBytes(familyNames[i]), Bytes.toBytes((String) fields.get(j)), valueBytes.get());
      }
    }

View Full Code Here


  @SuppressWarnings("unchecked")
  @Override
  public boolean source(FlowProcess<JobConf> flowProcess, SourceCall<Object[], RecordReader> sourceCall)
      throws IOException {
    Tuple result = new Tuple();


    Object key = sourceCall.getContext()[0];
    Object value = sourceCall.getContext()[1];
    boolean hasNext = sourceCall.getInput().next(key, value);
    if (!hasNext) {
      return false;
    }


    // Skip nulls
    if (key == null || value == null) {
      return true;
    }


    ImmutableBytesWritable keyWritable = (ImmutableBytesWritable) key;
    Result row = (Result) value;
    result.add(keyWritable);
    result.add(row);
    sourceCall.getIncomingEntry().setTuple(result);
    return true;
  }

View Full Code Here

  @SuppressWarnings("unchecked")
  @Override
  public void sink(FlowProcess<JobConf> flowProcess, SinkCall<Object[], OutputCollector> sinkCall) throws IOException {
    TupleEntry tupleEntry = sinkCall.getOutgoingEntry();
    OutputCollector outputCollector = sinkCall.getOutput();
    Tuple key = tupleEntry.selectTuple(RowKeyField);
    Object okey = key.getObject(0);
    ImmutableBytesWritable keyBytes = getBytes(okey);
    Put put = new Put(keyBytes.get());
    Fields outFields = tupleEntry.getFields().subtract(RowKeyField);
    if (null != outFields) {
      TupleEntry values = tupleEntry.selectEntry(outFields);

View Full Code Here

        boolean result = sourceCall.getInput().next( key, value );


        if( !result )
            return false;


        Tuple newTuple = ( (TupleRecord) value ).getTuple();
        sourceCall.getIncomingEntry().setTuple( newTuple );


        return true;
    }

View Full Code Here

        // it's ok to use NULL here so the collector does not write anything
        TupleEntry tupleEntry = sinkCall.getOutgoingEntry();
        OutputCollector outputCollector = sinkCall.getOutput();
        if( updateBy != null )
        {
            Tuple allValues = tupleEntry.selectTuple( updateValueFields );
            Tuple updateValues = tupleEntry.selectTuple( updateByFields );


            allValues = cleanTuple( allValues );


            TupleRecord key = new TupleRecord( allValues );


            if( updateValues.equals( updateIfTuple ) )
                outputCollector.collect( key, null );
            else
                outputCollector.collect( key, key );


            return;
        }


        Tuple result = tupleEntry.selectTuple( getSinkFields() );


        result = cleanTuple( result );


        outputCollector.collect( new TupleRecord( result ), null );
    }

View Full Code Here

        }
        
    }


    public void readFields( ResultSet resultSet ) throws SQLException {
        tuple = new Tuple();


        for( int i = 0; i < resultSet.getMetaData().getColumnCount(); i++ )
            tuple.add( (Comparable) resultSet.getObject( i + 1 ) );
    }

View Full Code Here

        boolean result = sourceCall.getInput().next(key, value);


        if (!result)
            return false;


        sourceCall.getIncomingEntry().setTuple(new Tuple(value.key, value.value));
        return true;
    }

View Full Code Here

        return true;
    }


    @Override public void sink(FlowProcess<JobConf> flowProcess,
        SinkCall<Object[], OutputCollector> sinkCall) throws IOException {
        Tuple tuple = sinkCall.getOutgoingEntry().getTuple();


        int shard = tuple.getInteger(0);
        Object f1 = tuple.getObject(1);
        Object f2 = tuple.getObject(2);
        
        byte[] key = (byte[]) f1;
        byte[] val = (byte[]) f2;
        KeyValDocument pair = new KeyValDocument(key, val);

View Full Code Here


        public void operate(FlowProcess process, FunctionCall call) {
            Object key = call.getArguments().getObject(0);


            int shard = shardIndex((byte[])key);
            call.getOutputCollector().add(new Tuple(shard));
        }

View Full Code Here


        public void operate(FlowProcess process, FunctionCall call) {
            Object f1 = call.getArguments().getObject(0);
            byte[] key = (byte[]) f1;
            BytesWritable sortField = new BytesWritable(key);
            call.getOutputCollector().add(new Tuple(sortField));
        }

View Full Code Here

0 1 2 3 4 5 6 7 8 9

TOP

Related Classes of cascading.tuple.Tuple

bixo.config.BixoPlatform

bixo.datum.FetchSetDatum

bixo.datum.ParsedDatum

bixo.examples.webmining.AnalyzedDatum

bixo.operations.FilterAndScoreByUrlAndRobotsTest

cascading.BasicPipesPlatformTest

cascading.BufferPipesPlatformTest

cascading.CoGroupFieldedPipesPlatformTest

cascading.DistanceUseCasePlatformTest

cascading.FieldedPipesPlatformTest

All source code are property of their respective owners. Java is a trademark of Sun Microsystems, Inc and owned by ORACLE Inc. Contact coftware#gmail.com.