Package org.apache.crunch

Examples of org.apache.crunch.Pipeline.readTextFile()

67
68
69
70
71
72
73
74
75
76
77
  @Test
  public void testAvro() throws Exception {
    Pipeline pipeline = new MRPipeline(AggregateIT.class, tmpDir.getDefaultConfiguration());
    String shakesInputPath = tmpDir.copyResourceFileName("shakes.txt");
    PCollection<String> shakes = pipeline.readTextFile(shakesInputPath);
    runMinMax(shakes, AvroTypeFamily.getInstance());
    pipeline.done();
  }

  @Test
View Full Code Here
110
111
112
113
114
115
116
117
118
119
120
  @Test
  public void testCollectUrls() throws Exception {
    Pipeline p = new MRPipeline(AggregateIT.class, tmpDir.getDefaultConfiguration());
    String urlsInputPath = tmpDir.copyResourceFileName("urls.txt");
    PTable<String, Collection<String>> urls = Aggregate.collectValues(p.readTextFile(urlsInputPath).parallelDo(
        new SplitFn(), tableOf(strings(), strings())));
    for (Pair<String, Collection<String>> e : urls.materialize()) {
      String key = e.first();
      int expectedSize = 0;
      if ("www.A.com".equals(key)) {
View Full Code Here
160
161
162
163
164
165
166
167
168
169
170
  }

  @Test
  public void testCollectValues_Writables() throws IOException {
    Pipeline pipeline = new MRPipeline(AggregateIT.class, tmpDir.getDefaultConfiguration());
    Map<Integer, Collection<Text>> collectionMap = pipeline.readTextFile(tmpDir.copyResourceFileName("set2.txt"))
        .parallelDo(new MapStringToTextPair(), Writables.tableOf(Writables.ints(), Writables.writables(Text.class)))
        .collectValues().materializeToMap();

    assertEquals(1, collectionMap.size());
View Full Code Here
174
175
176
177
178
179
180
181
182
183
184
  @Test
  public void testCollectValues_Avro() throws IOException {

    MapStringToEmployeePair mapFn = new MapStringToEmployeePair();
    Pipeline pipeline = new MRPipeline(AggregateIT.class, tmpDir.getDefaultConfiguration());
    Map<Integer, Collection<Employee>> collectionMap = pipeline.readTextFile(tmpDir.copyResourceFileName("set2.txt"))
        .parallelDo(mapFn, Avros.tableOf(Avros.ints(), Avros.records(Employee.class))).collectValues()
        .materializeToMap();

    assertEquals(1, collectionMap.size());
View Full Code Here
41
42
43
44
45
46
47
48
49
50
51
  };

  @Test
  public void testKill() throws Exception {
    Pipeline pipeline = new MRPipeline(FailIT.class, tempDir.getDefaultConfiguration());
    PCollection<String> p = pipeline.readTextFile(tempDir.copyResourceFileName("shakes.txt"));
    PCollection<Integer> result = p.parallelDo(new InverseFn(), Writables.ints());
    result.cache();

    PipelineExecution execution = pipeline.runAsync();
View Full Code Here
49
50
51
52
53
54
55
56
57
58
59
  public transient TemporaryPath tmpDir = TemporaryPaths.create();

  @Test
  public void testReflection() throws IOException {
    Pipeline pipeline = new MRPipeline(AvroReflectIT.class, tmpDir.getDefaultConfiguration());
    PCollection<StringWrapper> stringWrapperCollection = pipeline.readTextFile(tmpDir.copyResourceFileName("set1.txt"))
        .parallelDo(new MapFn<String, StringWrapper>() {

          @Override
          public StringWrapper map(String input) {
            StringWrapper stringWrapper = new StringWrapper();
View Full Code Here
75
76
77
78
79
80
81
82
83
84
85
  // doesn't crash
  @Test
  public void testCombinationOfReflectionAndSpecific() throws IOException {
    Assume.assumeTrue(Avros.CAN_COMBINE_SPECIFIC_AND_REFLECT_SCHEMAS);
    Pipeline pipeline = new MRPipeline(AvroReflectIT.class, tmpDir.getDefaultConfiguration());
    PCollection<Pair<StringWrapper, Person>> hybridPairCollection = pipeline.readTextFile(
        tmpDir.copyResourceFileName("set1.txt")).parallelDo(new MapFn<String, Pair<StringWrapper, Person>>() {

      @Override
      public Pair<StringWrapper, Person> map(String input) {
        Person person = new Person();
View Full Code Here
115
116
117
118
119
120
121
122
123
124
125
      Avros.reflects(StringWrapper.class));

  @Test
  public void testDerivedReflection() throws Exception {
    Pipeline pipeline = new MRPipeline(AvroReflectIT.class, tmpDir.getDefaultConfiguration());
    PCollection<String> stringWrapperCollection = pipeline.readTextFile(tmpDir.copyResourceFileName("set1.txt"))
        .parallelDo(IdentityFn.<String>getInstance(), STRING_PTYPE);
    List<String> strings = Lists.newArrayList(stringWrapperCollection.materialize());
    pipeline.done();
    assertEquals(Lists.newArrayList("b", "c", "a", "e"), strings);
  }
View Full Code Here
125
126
127
128
129
130
131
132
133
134
135
  }

  @Test
  public void testWrappedDerivedReflection() throws Exception {
    Pipeline pipeline = new MRPipeline(AvroReflectIT.class, tmpDir.getDefaultConfiguration());
    PCollection<Pair<Long, String>> stringWrapperCollection = pipeline.readTextFile(tmpDir.copyResourceFileName("set1.txt"))
        .parallelDo(new MapFn<String, Pair<Long, String>>() {
          @Override
          public Pair<Long, String> map(String input) {
            return Pair.of(1L, input);
          }
View Full Code Here
57
58
59
60
61
62
63
64
65
66
67
  @Test
  public void testWritables() throws Exception {
    Pipeline pipeline = new MRPipeline(AggregateIT.class, tmpDir.getDefaultConfiguration());
    String shakesInputPath = tmpDir.copyResourceFileName("shakes.txt");
    PCollection<String> shakes = pipeline.readTextFile(shakesInputPath);
    runMinMax(shakes, WritableTypeFamily.getInstance());
    pipeline.done();
  }

  @Test
View Full Code Here
TOP
Copyright © 2018 www.massapi.com. 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.