Java tutorial
/* * Cloud9: A MapReduce Library for Hadoop * * Licensed under the Apache License, Version 2.0 (the "License"); you * may not use this file except in compliance with the License. You may * obtain a copy of the License at * * http://www.apache.org/licenses/LICENSE-2.0 * * Unless required by applicable law or agreed to in writing, software * distributed under the License is distributed on an "AS IS" BASIS, * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or * implied. See the License for the specific language governing * permissions and limitations under the License. */ package edu.umd.cloud9.examples; import java.io.IOException; import java.util.Iterator; import java.util.StringTokenizer; //added Collections package import java.util.Collections; //added PairOfInts package import edu.umd.cloud9.io.PairOfInts; //added PairOfWritables package import edu.umd.cloud9.io.PairOfWritables; //added ArrayListWritable package import edu.umd.cloud9.io.ArrayListWritable; //added EntryObject2IntFrequencyDistribution package import edu.umd.cloud9.util.EntryObject2IntFrequencyDistribution; //added Object2IntFrequencyDistribution package import edu.umd.cloud9.util.Object2IntFrequencyDistribution; //added PairOfObjectInt package import edu.umd.cloud9.util.PairOfObjectInt; import org.apache.hadoop.conf.Configuration; import org.apache.hadoop.conf.Configured; import org.apache.hadoop.fs.FileSystem; import org.apache.hadoop.fs.Path; import org.apache.hadoop.io.IntWritable; import org.apache.hadoop.io.LongWritable; import org.apache.hadoop.io.Text; import org.apache.hadoop.mapred.FileInputFormat; import org.apache.hadoop.mapred.FileOutputFormat; import org.apache.hadoop.mapred.JobClient; import org.apache.hadoop.mapred.JobConf; import org.apache.hadoop.mapred.MapReduceBase; import org.apache.hadoop.mapred.Mapper; import org.apache.hadoop.mapred.OutputCollector; import org.apache.hadoop.mapred.Reducer; import org.apache.hadoop.mapred.Reporter; import org.apache.hadoop.util.Tool; import org.apache.hadoop.util.ToolRunner; import org.apache.log4j.Logger; /** * <p> * Simple inverted index. This Hadoop Tool builds an inverted index in flat text file, and * takes the following command-line arguments: * </p> * * <ul> * <li>[input-path] input path</li> * <li>[output-path] output path</li> * <li>[num-mappers] number of mappers</li> * <li>[num-reducers] number of reducers</li> * </ul> * * @author Jimmy Lin * @author Marc Sloan */ public class BuildInvertedIndex extends Configured implements Tool { private static final Logger sLogger = Logger.getLogger(BuildInvertedIndex.class); /** * Mapper: emits (term, term_freq) for every word occurrence * */ private static class MyMapper extends MapReduceBase implements Mapper<LongWritable, Text, Text, PairOfInts> { /** * Store an Object2IntFrequencyDistribution, which will be mapped * to each word found in the test */ //variable to hold counts private final static Object2IntFrequencyDistribution<String> counts = new EntryObject2IntFrequencyDistribution<String>(); /** * reuse objects to save overhead of object creation */ //variable to hold term private Text term = new Text(); /** * Mapping function. This takes the text input, converts it into a String which is split into * words, then each of the terms is mapped to the OutputCollector with a term frequency. * * @param docno Input key, not used in this example * @param doc A line of input Text taken from the data * @param output Map from each term (Text) to its term_freq (PairOfInts) */ public void map(LongWritable docno, Text doc, OutputCollector<Text, PairOfInts> output, Reporter reporter) throws IOException { //Convert input word into String and tokenize to find words String line = ((Text) doc).toString(); StringTokenizer itr = new StringTokenizer(line); //clear counts counts.clear(); //variable to hold term String word = null; //For each bigram, map it to a count of one. Duplicate bigrams will be counted //in the reduce phase. while (itr.hasMoreTokens()) { //update the word as the next token/word word = itr.nextToken(); //if word exists if (word != null) { //increment frequency of word counts.increment(word); } } //for every count in counts for (PairOfObjectInt<String> count : counts) { //set term as left element of count term.set(count.getLeftElement()); //set term frequency as the document number + the right element of count PairOfInts term_freq = new PairOfInts((int) docno.get(), count.getRightElement()); //output term and its frequency output.collect(term, term_freq); } } } /** * Reducer: sums up all the counts * */ private static class MyReducer extends MapReduceBase implements Reducer<Text, PairOfInts, Text, PairOfWritables<IntWritable, ArrayListWritable<PairOfInts>>> { /** * Stores the sum of counts for a term */ //variable to hold document frequency private final static IntWritable doc_freq = new IntWritable(); /** * @param docno The Text term * @param values An iterator over the values associated with this term * @param output Map from each docno (Text) to its inverted_index (PairOfWritables) * @param reporter Used to report progress */ public void reduce(Text docno, Iterator<PairOfInts> doc, OutputCollector<Text, PairOfWritables<IntWritable, ArrayListWritable<PairOfInts>>> output, Reporter reporter) throws IOException { //sum up values //variable to hold frequency int freq = 0; //variable to hold postings ArrayListWritable<PairOfInts> postings = new ArrayListWritable<PairOfInts>(); //until there are values left while (doc.hasNext()) { //add each value to the postings array by cloning it postings.add(doc.next().clone()); //increment frequency freq++; } //sort postings array in ascending order Collections.sort(postings); //set document frequency doc_freq.set(freq); //form the inverted index: document frequency + postings PairOfWritables<IntWritable, ArrayListWritable<PairOfInts>> inverted_index = new PairOfWritables<IntWritable, ArrayListWritable<PairOfInts>>( doc_freq, postings); //output document number + inverted index = docno, doc freq, postings output.collect(docno, inverted_index); } } /** * Creates an instance of this tool. */ public BuildInvertedIndex() { } /** * Prints argument options * @return */ private static int printUsage() { System.out.println("usage: [input-path] [output-path] [num-mappers] [num-reducers]"); ToolRunner.printGenericCommandUsage(System.out); return -1; } /** * Runs this tool. */ public int run(String[] args) throws Exception { if (args.length != 4) { printUsage(); return -1; } String inputPath = args[0]; String outputPath = args[1]; int mapTasks = Integer.parseInt(args[2]); int reduceTasks = Integer.parseInt(args[3]); sLogger.info("Tool: BuildInvertedIndex"); sLogger.info(" - input path: " + inputPath); sLogger.info(" - output path: " + outputPath); sLogger.info(" - number of mappers: " + mapTasks); sLogger.info(" - number of reducers: " + reduceTasks); JobConf conf = new JobConf(BuildInvertedIndex.class); conf.setJobName("BuildInvertedIndex"); conf.setNumMapTasks(mapTasks); conf.setNumReduceTasks(reduceTasks); FileInputFormat.setInputPaths(conf, new Path(inputPath)); FileOutputFormat.setOutputPath(conf, new Path(outputPath)); FileOutputFormat.setCompressOutput(conf, false); /** * Note that these must match the Class arguments given in the mapper */ conf.setOutputKeyClass(Text.class); conf.setOutputValueClass(PairOfInts.class); conf.setMapperClass(MyMapper.class); conf.setReducerClass(MyReducer.class); // Delete the output directory if it exists already Path outputDir = new Path(outputPath); FileSystem.get(outputDir.toUri(), conf).delete(outputDir, true); long startTime = System.currentTimeMillis(); JobClient.runJob(conf); sLogger.info("Job Finished in " + (System.currentTimeMillis() - startTime) / 1000.0 + " seconds"); return 0; } /** * Dispatches command-line arguments to the tool via the * <code>ToolRunner</code>. */ public static void main(String[] args) throws Exception { int res = ToolRunner.run(new Configuration(), new BuildInvertedIndex(), args); System.exit(res); } }