Package org.apache.hadoop.util.bloom

Examples of org.apache.hadoop.util.bloom.Key

234
235
236
237
238
239
240
241
      }
    }

    @Override
    public boolean accept(Pair<K, V> input) {
      Key key = new Key(keyToBytesFn.map(input.first()));
      return bloomFilter.membershipTest(key);
    }
View Full Code Here
101
102
103
104
105
106
107
108
   * Adds an int to the bloom filter.
   */
  public void addInt(int i) {
    ByteBuffer buf = ByteBuffer.allocate(4);
    buf.putInt(i);
    Key k = new Key(buf.array());
    bloom.add(k);
  }
View Full Code Here
136
137
138
139
140
141
142
143
144
145
146
          } catch (URISyntaxException e) {
          }
          builder = new StringBuilder();
          value = builder.append("<").append(value).append(">").toString(); //String builder
          uriSet.add(value);
          uris.add(new Key(value.getBytes(UTF8_CHARSET)));
        }
      }
    }
    Set<String> maps = new HashSet<String>();
    Map<String, String> bloomfilterMapping = new HashMap<String, String>();
View Full Code Here
127
128
129
130
131
132
133
134
135
136
137
            }
          } catch (URISyntaxException e) {
          }
          value = builder.append("<").append(value).append(">").toString(); //String builder
          uriSet.add(value);
          uris.add(new Key(value.getBytes(UTF8_CHARSET)));
        }
      }
    }
    Set<String> maps = new HashSet<String>();
    Map<String, String> bloomfilterMapping = new HashMap<String, String>();
View Full Code Here
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
    try {

      KR2RMLBloomFilterManager peopleBloomFilterManager = getBloomFilterManagerForSource("people.json", InputType.JSON, "people-model");
      KR2RMLBloomFilterManager scheduleBloomFilterManager = getBloomFilterManagerForSource("schedule.csv", InputType.CSV, "schedule-model");
      KR2RMLBloomFilter peoplePersonWithTwitterIdBF = peopleBloomFilterManager.getBloomFilter("http://isi.edu/integration/karma/dev#PredicateObjectMap_5fcf2d39-f62b-4cdd-863e-bde21493e1bd");
      Key k = new Key(("<http://lod.isi.edu/cs548/person/Slepicka>").getBytes());
      assertFalse(peoplePersonWithTwitterIdBF.membershipTest(k));
      k = new Key(("<http://lod.isi.edu/cs548/person/Taheriyan>").getBytes());
      assertTrue(peoplePersonWithTwitterIdBF.membershipTest(k));
      k = new Key(("<http://lod.isi.edu/cs548/person/Kozareva>").getBytes());
      assertFalse(peoplePersonWithTwitterIdBF.membershipTest(k));
      k = new Key(("<http://lod.isi.edu/cs548/person/Ambite>").getBytes());
      assertFalse(peoplePersonWithTwitterIdBF.membershipTest(k));
      k = new Key(("<http://lod.isi.edu/cs548/person/Szekely>").getBytes());
      assertTrue(peoplePersonWithTwitterIdBF.membershipTest(k));
      k = new Key(("<http://lod.isi.edu/cs548/person/Knoblock>").getBytes());
      assertTrue(peoplePersonWithTwitterIdBF.membershipTest(k));
      k = new Key(("<http://lod.isi.edu/cs548/person/Wu>").getBytes());
      assertFalse(peoplePersonWithTwitterIdBF.membershipTest(k));
      assertEquals(3, peoplePersonWithTwitterIdBF.estimateNumberOfHashedValues());
      KR2RMLBloomFilter schedulePersonBF = scheduleBloomFilterManager.getBloomFilter("http://isi.edu/integration/karma/dev#TriplesMap_413a6176-d893-45aa-b1c2-6661b5c491ab");
      k = new Key(("<http://lod.isi.edu/cs548/person/Slepicka>").getBytes());
      assertTrue(schedulePersonBF.membershipTest(k));
      k = new Key(("<http://lod.isi.edu/cs548/person/Taheriyan>").getBytes());
      assertTrue(schedulePersonBF.membershipTest(k));
      k = new Key(("<http://lod.isi.edu/cs548/person/Ambite>").getBytes());
      assertTrue(schedulePersonBF.membershipTest(k));
      k = new Key(("<http://lod.isi.edu/cs548/person/Szekely>").getBytes());
      assertTrue(schedulePersonBF.membershipTest(k));
      k = new Key(("<http://lod.isi.edu/cs548/person/Knoblock>").getBytes());
      assertTrue(schedulePersonBF.membershipTest(k));
      assertEquals(5, schedulePersonBF.estimateNumberOfHashedValues());
     
      KR2RMLBloomFilter intersectionBF = new KR2RMLBloomFilter(KR2RMLBloomFilter.defaultVectorSize, KR2RMLBloomFilter.defaultnbHash, Hash.JENKINS_HASH);
      intersectionBF.or(peoplePersonWithTwitterIdBF);
      intersectionBF.and(schedulePersonBF);
      assertEquals(3, intersectionBF.estimateNumberOfHashedValues());
     
      KR2RMLBloomFilter hasInstructorBF = scheduleBloomFilterManager.getBloomFilter("http://isi.edu/integration/karma/dev#RefObjectMap_bb82f923-2953-4bd4-bc7b-d1196e05dbf6");
     
      k = new Key(("<http://lod.isi.edu/cs548/person/Szekely>").getBytes());
      assertTrue(hasInstructorBF.membershipTest(k));
      intersectionBF = new KR2RMLBloomFilter(KR2RMLBloomFilter.defaultVectorSize, KR2RMLBloomFilter.defaultnbHash, Hash.JENKINS_HASH);
      intersectionBF.or(hasInstructorBF);
      intersectionBF.and(peoplePersonWithTwitterIdBF);
      assertEquals(3, intersectionBF.estimateNumberOfHashedValues());
View Full Code Here
50
51
52
53
54
55
56
57
58
    {
      idToBloomFilter.putIfAbsent(id, new KR2RMLBloomFilter(KR2RMLBloomFilter.defaultVectorSize, KR2RMLBloomFilter.defaultnbHash, Hash.JENKINS_HASH));
    }
    bf = idToBloomFilter.get(id);
   
    Key k = new Key(uri.getBytes(UTF8_CHARSET));
    bf.add(k);
    return;
  }
View Full Code Here
82
83
84
85
86
87
88
89
90
91
92
93
94
95
      Statement st = iterator.next();
      String subject = "<" + st.getSubject().toString() + ">";
      String object = "<" + st.getObject().toString() + ">";
      for (Entry<String, KR2RMLBloomFilter> entry : mapping.entrySet()) {
        KR2RMLBloomFilter bf = entry.getValue();
        if (bf.membershipTest(new Key(subject.getBytes("UTF-8"))))
          bf.add(new Key(object.getBytes("UTF-8")));
        if (bf.membershipTest(new Key(object.getBytes("UTF-8"))))
          bf.add(new Key(subject.getBytes("UTF-8")));
      }
    }
   
    utilObj.updateTripleStoreWithBloomFilters(mapping, serializedmapping, triplestoreURL, null);
   
View Full Code Here
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
      @Override
      public Collection<Key> generateKeys(String input) {
        List<String> parts = Arrays.asList(StringUtils.split(input, " "));
        Collection<Key> keys = new HashSet<Key>();
        for (String stringpart : parts) {
          keys.add(new Key(stringpart.getBytes()));
        }
        return keys;
      }
    };
    Map<String, BloomFilter> filterValues = BloomFilterFactory.createFilter(new Path(inputPath), filterFn).getValue();
    assertEquals(1, filterValues.size());
    BloomFilter filter = filterValues.get("shakes.txt");
    assertTrue(filter.membershipTest(new Key("Mcbeth".getBytes())));
    assertTrue(filter.membershipTest(new Key("apples".getBytes())));
  }
View Full Code Here
83
84
85
86
87
88
89
90
    public void close() throws IOException {
      if (collector!=null) {
        collector.collect(nullKey, bloomFilter);
      }
      System.err.println("TESTTTTTTTTTTTTTTT [organizations elsewhere] "+
          bloomFilter.membershipTest(new Key("organizations elsewhere".getBytes())));
      super.close();
    }
View Full Code Here
99
100
101
102
103
104
105
        this.collector = collector;
      }     
          
      byte[] buff = new byte[key.getLength()];
      System.arraycopy(key.getBytes(), 0, buff, 0, buff.length);
      bloomFilter.add(new Key(buff, value.get()));
    }
View Full Code Here
TOP

Related Classes of org.apache.hadoop.util.bloom.Key

  • brickhouse.udf.bloom.BloomContainsUDF
  • brickhouse.udf.bloom.BloomTest
  • brickhouse.udf.bloom.BloomUDAF$BloomUDAFEvaluator
  • cc.bf.NgramsToBloomFilter$BFBase
  • cc.bf.NgramsToBloomFilter$BFMapper
  • cc.bf.ReadNgram
  • com.cloudera.util.bloom.BloomSet
  • com.ontology2.bakemono.bloom.BloomReducer
  • com.ontology2.bakemono.bloom.BloomReducerTest
  • edu.isi.karma.controller.command.alignment.SearchForDataToAugmentCommand
    • 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.