Package org.apache.crunch

Examples of org.apache.crunch.Pipeline

32
33
34
35
36
37
38
39
40
  public transient TemporaryPath tmpDir = TemporaryPaths.create();
 
  @Test(expected=CrunchRuntimeException.class)
  public void testTextToAvro() throws Exception {
    String shakes = tmpDir.copyResourceFileName("shakes.txt");
    Pipeline pipeline = new MRPipeline(TextToAvroIT.class, tmpDir.getDefaultConfiguration());
    pipeline.read(From.textFile(shakes)).write(To.avroFile("output"));
    pipeline.run();
  }
View Full Code Here
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
    new File(infilename).getParentFile().mkdirs();

    writeFile(expected, infilename);

    Pipeline memPipeline = MemPipeline.getInstance();
    PCollection<String> memPColl = memPipeline.readTextFile(infilename);
    Target memTarget = new TextFileTarget(memOutFilename);
    memPipeline.write(memPColl, memTarget, WriteMode.OVERWRITE);
    memPipeline.run();
    File outDir = new File(memOutFilename);
    File actualMemOut = null;
    for (File f : outDir.listFiles()) {
      String name = f.getName();
      if (name.contains("out") && name.endsWith(".txt")) {
        actualMemOut = f;
        break;
      }
    }
    String actualMemText = Files.readFirstLine(actualMemOut, Charsets.UTF_8);

    Pipeline mrPipeline = new MRPipeline(getClass());
    PCollection<String> mrPColl = mrPipeline.readTextFile(infilename);
    Target mrTarget = new TextFileTarget(mrOutFilename);
    mrPipeline.write(mrPColl, mrTarget, WriteMode.OVERWRITE);
    mrPipeline.run();
    String actualMrText = Files.readFirstLine(new File(mrOutFilename + "/part-m-00000"), Charsets.UTF_8);

    Assert.assertEquals("MR file mismatch", expected, actualMrText);
    Assert.assertEquals("Mem file mismatch", expected, actualMemText);
  }
View Full Code Here
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
    }
  }

  @Test
  public void testMapper() throws Exception {
    Pipeline p = new MRPipeline(MapreduceIT.class, tempDir.getDefaultConfiguration());
    Path shakesPath = tempDir.copyResourcePath("shakes.txt");
    PCollection<String> in = p.read(From.textFile(shakesPath));
    PTable<IntWritable, Text> two = in.parallelDo(new MapFn<String, Pair<IntWritable, Text>>() {
      @Override
      public Pair<IntWritable, Text> map(String input) {
        return Pair.of(new IntWritable(input.length()), new Text(input));
      }
    }, Writables.tableOf(Writables.writables(IntWritable.class), Writables.writables(Text.class)));
   
    PTable<IntWritable, Text> out = Mapreduce.map(two, TestMapper.class, IntWritable.class, Text.class);
    out.write(To.sequenceFile(tempDir.getPath("temp")));
    PipelineResult res = p.done();
    assertEquals(1, res.getStageResults().size());
    StageResult sr = res.getStageResults().get(0);
    assertEquals(3667, sr.getCounters().findCounter("written", "out").getValue());
  }
View Full Code Here
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
    assertEquals(3667, sr.getCounters().findCounter("written", "out").getValue());
  }
 
  @Test
  public void testReducer() throws Exception {
    Pipeline p = new MRPipeline(MapredIT.class, tempDir.getDefaultConfiguration());
    Path shakesPath = tempDir.copyResourcePath("shakes.txt");
    PCollection<String> in = p.read(From.textFile(shakesPath));
    PTable<IntWritable, Text> two = in.parallelDo(new MapFn<String, Pair<IntWritable, Text>>() {
      @Override
      public Pair<IntWritable, Text> map(String input) {
        return Pair.of(new IntWritable(input.length()), new Text(input));
      }
    }, Writables.tableOf(Writables.writables(IntWritable.class), Writables.writables(Text.class)));
   
    PTable<Text, LongWritable> out = Mapreduce.reduce(two.groupByKey(), TestReducer.class, Text.class, LongWritable.class);
    out.write(To.sequenceFile(tempDir.getPath("temp")));
    PipelineResult res = p.done();
    assertEquals(1, res.getStageResults().size());
    StageResult sr = res.getStageResults().get(0);
    assertEquals(19, sr.getCounters().findCounter("words", "where").getValue());
  }
View Full Code Here
96
97
98
99
100
101
102
103
104
105
106
107
108
109
  }

  @Test
  public void testMemPipelineFileWriter() throws Exception {
    File outputDir = baseTmpDir.getFile("mempipe");
    Pipeline p = MemPipeline.getInstance();
    PCollection<String> lines = MemPipeline.collectionOf("hello", "world");
    p.writeTextFile(lines, outputDir.toString());
    p.done();
    File outputFile = getOutputFile(outputDir, "*.txt");

    List<String> txt = Files.readLines(outputFile, Charsets.UTF_8);
    assertEquals(ImmutableList.of("hello", "world"), txt);
  }
View Full Code Here
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
    }
  }
 
  @Test
  public void testMapper() throws Exception {
    Pipeline p = new MRPipeline(MapredIT.class, tempDir.getDefaultConfiguration());
    Path shakesPath = tempDir.copyResourcePath("shakes.txt");
    PCollection<String> in = p.read(From.textFile(shakesPath));
    PTable<IntWritable, Text> two = in.parallelDo(new MapFn<String, Pair<IntWritable, Text>>() {
      @Override
      public Pair<IntWritable, Text> map(String input) {
        return Pair.of(new IntWritable(input.length()), new Text(input));
      }
    }, Writables.tableOf(Writables.writables(IntWritable.class), Writables.writables(Text.class)));
   
    PTable<Text, LongWritable> out = Mapred.map(two, TestMapper.class, Text.class, LongWritable.class);
    out.write(To.sequenceFile(tempDir.getPath("temp")));
    PipelineResult res = p.done();
    assertEquals(1, res.getStageResults().size());
    StageResult sr = res.getStageResults().get(0);
    assertEquals(3667, sr.getCounters().findCounter("written", "out").getValue());
  }
View Full Code Here
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
    assertEquals(3667, sr.getCounters().findCounter("written", "out").getValue());
  }
 
  @Test
  public void testReducer() throws Exception {
    Pipeline p = new MRPipeline(MapredIT.class, tempDir.getDefaultConfiguration());
    Path shakesPath = tempDir.copyResourcePath("shakes.txt");
    PCollection<String> in = p.read(From.textFile(shakesPath));
    PTable<IntWritable, Text> two = in.parallelDo(new MapFn<String, Pair<IntWritable, Text>>() {
      @Override
      public Pair<IntWritable, Text> map(String input) {
        return Pair.of(new IntWritable(input.length()), new Text(input));
      }
    }, Writables.tableOf(Writables.writables(IntWritable.class), Writables.writables(Text.class)));
   
    PTable<Text, LongWritable> out = Mapred.reduce(two.groupByKey(), TestReducer.class, Text.class, LongWritable.class);
    out.write(To.sequenceFile(tempDir.getPath("temp")));
    PipelineResult res = p.done();
    assertEquals(1, res.getStageResults().size());
    StageResult sr = res.getStageResults().get(0);
    assertEquals(108, sr.getCounters().findCounter("thou", "count").getValue());
  }
View Full Code Here
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
public class CountersTest implements Serializable {

  @Test
  public void counterTest() throws Exception {
    Pipeline pipeline = MemPipeline.getInstance();

    // Single row PCollection.
    PCollection<String> objects = MemPipeline.collectionOf(Arrays.asList(new String[]{"hello world"}));
    System.out.println("Objects: " + ((MemCollection) objects).getCollection());

    // Counter creating Map.
    PCollection<String> objects2 = objects.parallelDo("Create counters",
        new MapFn<String, String>() {
          @Override
          public String map(String input) {
            for(int i = 0; i < 200; ++i) {
              this.increment("testCounter", String.valueOf(i));
            }
            return input;
          }
        },
        Writables.strings()
    );

    // Run it!
    pipeline.done();
    System.out.println("Objects2: " + ((MemCollection) objects2).getCollection());
  }
View Full Code Here
39
40
41
42
43
44
45
46
47
48
49
50
  @Test
  public void testVanillaCSV() throws Exception {
    final String[] expectedFileContents = { "1,2,3,4", "5,6,7,8", "9,10,11", "12,13,14" };

    final String vanillaCSVFile = tmpDir.copyResourceFileName("vanilla.csv");
    final Pipeline pipeline = new MRPipeline(CSVFileSourceIT.class, tmpDir.getDefaultConfiguration());
    final PCollection<String> csvLines = pipeline.read(new CSVFileSource(new Path(vanillaCSVFile)));

    final Collection<String> csvLinesList = csvLines.asCollection().getValue();

    for (int i = 0; i < expectedFileContents.length; i++) {
      assertTrue(csvLinesList.contains(expectedFileContents[i]));
View Full Code Here
40
41
42
43
44
45
46
47
48
49
50
51
  @Test
  public void testShard() throws Exception {
    File inDir = tempDir.getFile("in");
    FileUtils.writeLines(new File(inDir, "part1"), ImmutableList.of("part1", "part1"));
    FileUtils.writeLines(new File(inDir, "part2"), ImmutableList.of("part2"));
    Pipeline pipeline = new MRPipeline(ShardIT.class);
    PCollection<String> in = pipeline.read(From.textFile(inDir.getPath()));
    // We can only test on 1 shard here, as local MR does not support multiple reducers.
    PCollection<String> out = Shard.shard(in, 1);
    assertEquals(
        ImmutableMultiset.copyOf(out.materialize()),
        ImmutableMultiset.of("part1", "part1", "part2"));
View Full Code Here
TOP

Related Classes of org.apache.crunch.Pipeline

  • $.WordCount
  • com.cloudera.cdk.data.crunch.TestCrunchDatasets
  • org.apache.bigtop.bigpetstore.etl.CrunchETL
  • org.apache.crunch.contrib.io.jdbc.DataBaseSourceIT
  • org.apache.crunch.examples.AverageBytesByIP
  • org.apache.crunch.examples.SecondarySortExample
  • org.apache.crunch.examples.TotalBytesByIP
  • org.apache.crunch.examples.TotalWordCount
  • org.apache.crunch.examples.WordAggregationHBase
  • org.apache.crunch.examples.WordCount
    • Copyright © 2018 www.massapicom. All rights reserved.
    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.