Hadoop Pipes Example

::::::::::::::
dp_hpt_mr.hh
::::::::::::::
#ifndef __DP_HPT_MR_HH__
#define __DP_HPT_MR_HH__

#include <Pipes.hh>

namespace dp_hpt
{
  class word_count_mapper : public HadoopPipes::Mapper
  {
  public:
    word_count_mapper(const HadoopPipes::TaskContext &) {};

    void map(HadoopPipes::MapContext & context);
  };

  class word_count_reducer : public HadoopPipes::Reducer
  {
  public:
    word_count_reducer(const HadoopPipes::TaskContext &) {};

    void reduce(HadoopPipes::ReduceContext & context);
  };
};

#endif //__DP_HPT_MR_HH__
::::::::::::::
dp_hadoop_pipes_test.cpp
::::::::::::::
#include "dp_hpt_mr.hh"

#include <iostream>
#include <TemplateFactory.hh>

using namespace std;

int
main(int argc,
     char * argv[])
{
  HadoopPipes::TemplateFactory2<dp_hpt::word_count_mapper,
                                dp_hpt::word_count_reducer> factory;

  return HadoopPipes::runTask(factory);
}
::::::::::::::
dp_hpt_mr.cpp
::::::::::::::
#include "dp_hpt_mr.hh"
#include <boost/algorithm/string.hpp>
#include <boost/tokenizer.hpp>
#include <boost/lexical_cast.hpp>
#include <string>

namespace dp_hpt
{
  void word_count_mapper::map(HadoopPipes::MapContext & context)
  {
    typedef boost::tokenizer<> tokenizer_t;
    tokenizer_t tokenizer(context.getInputValue());

    for(tokenizer_t::const_iterator i = tokenizer.begin();
	tokenizer.end() != i;
	++i)
      {
	context.emit(boost::to_lower_copy(*i), "1");
      }
  }

  void word_count_reducer::reduce(HadoopPipes::ReduceContext & context)
  {
    uint32_t count(0);

    do
      {
	++count;
      }
    while(context.nextValue());

    context.emit(context.getInputKey(),
		 boost::lexical_cast<std::string>(count));
  }
};
::::::::::::::
../conf/dp_hpt_conf.xml
::::::::::::::
<?xml version="1.0"?>
<configuration>
  <property>
    <name>hadoop.pipes.executable</name>
    <value>/path/to/dp_hadoop_pipes_test</value>
  </property>
  <property>
    <name>hadoop.pipes.java.recordreader</name>
    <value>true</value>
  </property>
  <property>
    <name>hadoop.pipes.java.recordwriter</name>
    <value>true</value>
  </property>
</configuration>