List of usage examples for org.apache.hadoop.io Text Text
public Text(byte[] utf8)
From source file:TestStringRelevance.java
License:Apache License
@Override public void setUp() throws Exception { fs.delete(new Path(INPUT), true); fs.delete(new Path(QUERY), true); fs.delete(new Path(OUTPUT), true); inputTap = new Hfs(new SequenceFile(new Fields("str1", "str2")), INPUT); TapCollector coll = new TapCollector(inputTap, new JobConf()); coll.add(tuple1);/*from ww w .ja va 2s . co m*/ coll.add(tuple2); coll.add(tuple3); coll.add(tuple4); coll.add(tuple5); coll.add(tuple6); coll.add(tuple7); coll.add(tuple8); coll.add(tuple9); coll.close(); keyTap = new Hfs(new SequenceFile(new Fields("str")), QUERY); coll = new TapCollector(keyTap, new JobConf()); coll.add(new Tuple(new Text("nathan@rapleaf.com"))); coll.add(new Tuple(new Text("1@gmail.com"))); coll.add(new Tuple(new Text("2@gmail.com"))); coll.add(new Tuple(new Text("6@gmail.com"))); coll.close(); outputTap = new Hfs(new SequenceFile(new Fields("str1", "str2")), OUTPUT); }
From source file:ZipFileRecordReader.java
License:Apache License
/** * This is where the magic happens, each ZipEntry is decompressed and * readied for the Mapper. The contents of each file is held *in memory* * in a BytesWritable object./*from w ww. j av a 2 s . c o m*/ * * If the ZipFileInputFormat has been set to Lenient (not the default), * certain exceptions will be gracefully ignored to prevent a larger job * from failing. */ @Override public boolean nextKeyValue() throws IOException, InterruptedException { ZipEntry entry = null; try { entry = zip.getNextEntry(); } catch (ZipException e) { if (ZipFileInputFormat.getLenient() == false) throw e; } // Sanity check if (entry == null) { isFinished = true; return false; } // Filename currentKey = new Text(entry.getName()); // Read the file contents ByteArrayOutputStream bos = new ByteArrayOutputStream(); byte[] temp = new byte[8192]; while (true) { int bytesRead = 0; try { bytesRead = zip.read(temp, 0, 8192); } catch (EOFException e) { if (ZipFileInputFormat.getLenient() == false) throw e; return false; } if (bytesRead > 0) bos.write(temp, 0, bytesRead); else break; } zip.closeEntry(); // Uncompressed contents currentValue = new BytesWritable(bos.toByteArray()); return true; }
From source file:DisplayClustering.java
License:Apache License
protected static void writeSampleData(Path output) throws IOException { Configuration conf = new Configuration(); FileSystem fs = FileSystem.get(output.toUri(), conf); SequenceFile.Writer writer = new SequenceFile.Writer(fs, conf, output, Text.class, VectorWritable.class); try {//from ww w . j ava2s.c o m int i = 0; for (VectorWritable vw : SAMPLE_DATA) { writer.append(new Text("sample_" + i++), vw); } } finally { Closeables.close(writer, false); } }
From source file:TestString.java
License:Apache License
@Test public void testStringSubstring() throws Exception { Text text = new Text("string"); Text text1 = new Text(); Text text2 = new Text(); long start = System.nanoTime(); for (int i = 0; i < 100000000; i++) { String str = text.toString(); String str1 = str.substring(0, 2); String str2 = str.substring(3, str.length()); text1.set(str1);/*from w ww.jav a2 s . c o m*/ text2.set(str2); } long end = System.nanoTime(); System.out.println("TextStringSubString"); System.out.println("text1: " + text1.toString()); System.out.println("text2: " + text2.toString()); System.out.println("Elapsed Time: " + (end - start) / 1000000000f + " seconds."); }
From source file:TestString.java
License:Apache License
@Test public void testTextSubstring() throws Exception { Text text = new Text("string"); Text text1 = new Text(); Text text2 = new Text(); long start = System.nanoTime(); for (int i = 0; i < 100000000; i++) { text1.set(text.getBytes(), 0, 2); text2.set(text.getBytes(), 3, text.getLength() - 3); }/* w w w . j a v a 2 s . c o m*/ long end = System.nanoTime(); System.out.println("TestTextSubString"); System.out.println("text1: " + text1.toString()); System.out.println("text2: " + text2.toString()); System.out.println("Elapsed Time: " + (end - start) / 1000000000f + " seconds."); }
From source file:Deducer.java
License:Apache License
public void reduce(Text key, Iterator<Text> values, OutputCollector<Text, Text> output, Reporter reporter) throws IOException { Set<String> attackers = new TreeSet<String>(); while (values.hasNext()) { String valStr = values.next().toString(); attackers.add(valStr);/*w w w.j av a 2s . c o m*/ } output.collect(key, new Text(attackers.toString())); }
From source file:PageRankIterationReducerTest.java
public void test1() { Text key = new Text("testKey"); PageRankFollower pageRankFollower = new PageRankFollower(new Text(key.toString()), 1.0, 0, new ArrayList<Text>()); List<PageRankFollower> values = new ArrayList<PageRankFollower>(); values.add(pageRankFollower);//ww w.j a v a2 s . c o m PageRankFollower result = (new PageRankIterationReducer()).calculatePageRank(key, values); System.out.println("result: " + result); }
From source file:TestBAM.java
License:Open Source License
@Override protected void map(LongWritable ignored, SAMRecordWritable wrec, org.apache.hadoop.mapreduce.Mapper<LongWritable, SAMRecordWritable, Text, SAMRecordWritable>.Context ctx) throws InterruptedException, IOException { final SAMRecord record = wrec.get(); int len = record.getInferredInsertSize();//record.getAlignmentEnd() - record.getAlignmentStart(); int chr = Integer.parseInt(record.getReferenceName()); if (Math.abs(len) > 1000 && chr == 20) { System.out.println(record.toString()); System.out.println("AbsLen:" + Integer.toString(Math.abs(len))); ctx.write(new Text(wrec.get().getReadName()), wrec); }/*from w w w . jav a2s .c om*/ }
From source file:TaggedMapOutput.java
License:Apache License
public TaggedMapOutput() { this.tag = new Text(""); }
From source file:SampleUdf.java
License:Apache License
public Text evaluate(final Text s, Text sleepTime) throws InterruptedException { Long time = 180 * 1000L;//from w w w. j a va2 s.c om if (sleepTime != null) { time = Long.parseLong(sleepTime.toString()) * 1000L; } System.out.println("Sleep Time : " + time); Thread.sleep(time); if (s == null) { return null; } return new Text(s.toString().toLowerCase()); }