org.apache.nutch.util.domain
Class DomainStatistics

java.lang.Object
  org.apache.hadoop.mapred.MapReduceBase
      org.apache.nutch.util.domain.DomainStatistics

All Implemented Interfaces:

Closeable, Configurable, JobConfigurable, Mapper<Text,CrawlDatum,Text,LongWritable>, Reducer<Text,LongWritable,LongWritable,Text>, Tool

public class DomainStatistics
extends MapReduceBase
implements Tool, Mapper<Text,CrawlDatum,Text,LongWritable>, Reducer<Text,LongWritable,LongWritable,Text>

Extracts some very basic statistics about domains from the crawldb

Nested Class Summary

static class	DomainStatistics.DomainStatisticsCombiner
static class	DomainStatistics.MyCounter

Constructor Summary

DomainStatistics()

Method Summary

void	configure(JobConf job)
Configuration	getConf()
static void	main(String[] args)
void	map(Text urlText, CrawlDatum datum, OutputCollector<Text,LongWritable> output, Reporter reporter)
void	reduce(Text key, Iterator<LongWritable> values, OutputCollector<LongWritable,Text> output, Reporter reporter)
int	run(String[] args)
void	setConf(Configuration conf)

Methods inherited from class org.apache.hadoop.mapred.MapReduceBase

close

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Methods inherited from interface java.io.Closeable

close

Constructor Detail

DomainStatistics

public DomainStatistics()

Method Detail

run

public int run(String[] args)
        throws IOException

Specified by:

run in interface Tool

Throws:

IOException

configure

public void configure(JobConf job)

Specified by:

configure in interface JobConfigurable

Overrides:

configure in class MapReduceBase

getConf

public Configuration getConf()

Specified by:

getConf in interface Configurable

setConf

public void setConf(Configuration conf)

Specified by:

setConf in interface Configurable

map

public void map(Text urlText,
                CrawlDatum datum,
                OutputCollector<Text,LongWritable> output,
                Reporter reporter)
         throws IOException

Specified by:

map in interface Mapper<Text,CrawlDatum,Text,LongWritable>

Throws:

IOException

reduce

public void reduce(Text key,
                   Iterator<LongWritable> values,
                   OutputCollector<LongWritable,Text> output,
                   Reporter reporter)
            throws IOException

Specified by:

reduce in interface Reducer<Text,LongWritable,LongWritable,Text>

Throws:

IOException

main

public static void main(String[] args)
                 throws Exception

Throws:

Exception