Package edu.umd.cloud9.collection.clue

Examples of edu.umd.cloud9.collection.clue.ClueWarcRecord

182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
          "-output_path=" + indexRootPath + "/wiki-docid-tmp",
          "-output_file=" + mappingFile.toString(),
          "-lang=" + collectionLang };
      LOG.info("Running BuildWikipediaDocnoMapping with args " + Arrays.toString(arr));

      BuildWikipediaDocnoMapping tool = new BuildWikipediaDocnoMapping();
      tool.setConf(conf);
      tool.run(arr);

      fs.delete(new Path(indexRootPath + "/wiki-docid-tmp"), true);
    }

    // Repack Wikipedia into sequential compressed block
    p = new Path(seqCollection);
    LOG.info(seqCollection + " doesn't exist, creating...");
    String[] arr = new String[] { "-input=" + rawCollection,
        "-output=" + seqCollection,
        "-mapping_file=" + mappingFile.toString(),
        "-compression_type=block",
        "-wiki_language=" + collectionLang };
    LOG.info("Running RepackWikipedia with args " + Arrays.toString(arr));

    RepackWikipedia tool = new RepackWikipedia();
    tool.setConf(conf);
    tool.run(arr);

    conf.set(Constants.CollectionName, "Wikipedia-"+collectionLang);
    conf.setInt(Constants.NumMapTasks, numMappers);
    conf.setInt(Constants.NumReduceTasks, numReducers);
    conf.set(Constants.CollectionPath, seqCollection);
View Full Code Here
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
    Path mappingFile = env.getDocnoMappingData();
    if (!fs.exists(mappingFile)) {
      LOG.info(mappingFile + " doesn't exist, creating...");
      String[] arr = new String[] { rawCollection, indexRootPath + "/wiki-docid-tmp",
          mappingFile.toString(), new Integer(numMappers).toString() };
      BuildWikipediaDocnoMapping tool = new BuildWikipediaDocnoMapping();
      tool.setConf(conf);
      tool.run(arr);

      fs.delete(new Path(indexRootPath + "/wiki-docid-tmp"), true);
    }else{
      LOG.info(p+" exists");
    }

    // Repack Wikipedia into sequential compressed block
    p = new Path(seqCollection);
    if (!fs.exists(p)) {
      LOG.info(seqCollection + " doesn't exist, creating...");
      String[] arr = new String[] { rawCollection, seqCollection, mappingFile.toString(), "block"};
      RepackWikipedia tool = new RepackWikipedia();
      tool.setConf(conf);
      tool.run(arr);
    }

    conf.set("Ivory.CollectionName", "Wikipedia-"+collectionLang);
    conf.setInt("Ivory.NumMapTasks", numMappers);
    conf.setInt("Ivory.NumReduceTasks", numReducers);
View Full Code Here
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
          "-output_path=" + indexRootPath + "/wiki-docid-tmp",
          "-output_file=" + mappingFile.toString(),
          "-wiki_language=" + collectionLang };
      LOG.info("Running BuildWikipediaDocnoMapping with args " + Arrays.toString(arr));

      BuildWikipediaDocnoMapping tool = new BuildWikipediaDocnoMapping();
      tool.setConf(conf);
      tool.run(arr);

      fs.delete(new Path(indexRootPath + "/wiki-docid-tmp"), true);
    } else {
      LOG.info("Docno mapping already exists at: " + mappingFile);
    }

    // Repack Wikipedia into sequential compressed block
    if (!fs.exists(new Path(seqCollection + "/part-00000"))) {
      LOG.info(seqCollection + " doesn't exist, creating...");
      String[] arr = new String[] { "-input=" + rawCollection,
          "-output=" + seqCollection,
          "-mapping_file=" + mappingFile.toString(),
          "-compression_type=block",
          "-wiki_language=" + collectionLang };
      LOG.info("Running RepackWikipedia with args " + Arrays.toString(arr));

      RepackWikipedia tool = new RepackWikipedia();
      tool.setConf(conf);
      tool.run(arr);
    } else {
      LOG.info("Repacked collection already exists at: " + seqCollection);     
    }

    conf.set(Constants.CollectionName, "Wikipedia-"+collectionLang);
View Full Code Here
199
200
201
202
203
204
205
206
207
208
209
210
211
        "-mapping_file=" + mappingFile.toString(),
        "-compression_type=block",
        "-wiki_language=" + collectionLang };
    LOG.info("Running RepackWikipedia with args " + Arrays.toString(arr));

    RepackWikipedia tool = new RepackWikipedia();
    tool.setConf(conf);
    tool.run(arr);

    conf.set(Constants.CollectionName, "Wikipedia-"+collectionLang);
    conf.setInt(Constants.NumMapTasks, numMappers);
    conf.setInt(Constants.NumReduceTasks, numReducers);
    conf.set(Constants.CollectionPath, seqCollection);
View Full Code Here
167
168
169
170
171
172
173
174
175
176
177
178
179
    // Repack Wikipedia into sequential compressed block
    p = new Path(seqCollection);
    if (!fs.exists(p)) {
      LOG.info(seqCollection + " doesn't exist, creating...");
      String[] arr = new String[] { rawCollection, seqCollection, mappingFile.toString(), "block"};
      RepackWikipedia tool = new RepackWikipedia();
      tool.setConf(conf);
      tool.run(arr);
    }

    conf.set("Ivory.CollectionName", "Wikipedia-"+collectionLang);
    conf.setInt("Ivory.NumMapTasks", numMappers);
    conf.setInt("Ivory.NumReduceTasks", numReducers);
View Full Code Here
210
211
212
213
214
215
216
217
218
219
220
221
222
          "-mapping_file=" + mappingFile.toString(),
          "-compression_type=block",
          "-wiki_language=" + collectionLang };
      LOG.info("Running RepackWikipedia with args " + Arrays.toString(arr));

      RepackWikipedia tool = new RepackWikipedia();
      tool.setConf(conf);
      tool.run(arr);
    } else {
      LOG.info("Repacked collection already exists at: " + seqCollection);     
    }

    conf.set(Constants.CollectionName, "Wikipedia-"+collectionLang);
View Full Code Here
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
        if (fileStats[i].getPath().getName().startsWith("_")) {
          continue;
        }

        LOG.info("processing " + fileStats[i].getPath());
        FSLineReader reader = new FSLineReader(fileStats[i].getPath(), fs);

        Text line = new Text();
        while (reader.readLine(line) > 0) {
          String[] arr = line.toString().split("\\t+", 2);

          int docno = Integer.parseInt(arr[0]);
          int len = Integer.parseInt(arr[1]);

          // Note that because of speculative execution there may be
          // multiple copies of doclength data. Therefore, we can't
          // just count number of doclengths read. Instead, keep track
          // of largest docno encountered.
          if (docno < docnoOffset) {
            throw new RuntimeException(
                "Error: docno " + docno + " < docnoOffset " + docnoOffset + "!");
          }

          doclengths[docno - docnoOffset] = len;

          if (docno > maxDocno) {
            maxDocno = docno;
          }
          if (docno < minDocno) {
            minDocno = docno;
          }
        }
        reader.close();
        context.getCounter(DocLengths.Files).increment(1);
      }

      LOG.info("min docno: " + minDocno);
      LOG.info("max docno: " + maxDocno);
View Full Code Here
23
24
25
26
27
28
29
   * @param size
   *     number of bits
   */
  public MinhashSignature(){
    super();
    terms = new ArrayListOfIntsWritable();
  }
View Full Code Here
27
28
29
30
31
32
33
    super();
    terms = new ArrayListOfIntsWritable();
  }

  public MinhashSignature(ArrayListOfIntsWritable b){
    terms = new ArrayListOfIntsWritable(b);
  }
View Full Code Here
36
37
38
39
40
41
42
    this(other.terms);
  }

  public MinhashSignature(int numTerms){      //need this constructor for general purposes.
    super();
    terms = new ArrayListOfIntsWritable(numTerms);
  }
View Full Code Here
TOP

Related Classes of edu.umd.cloud9.collection.clue.ClueWarcRecord

  • bak.pcj.IntIterator
  • edu.umd.cloud9.example.bfs.BfsNodeTest
  • edu.umd.cloud9.example.pagerank.PageRankNodeTest
  • edu.umd.cloud9.io.array.ArrayListOfIntsWritable
  • edu.umd.cloud9.io.fastutil.Int2FloatOpenHashMapWritableTest
  • edu.umd.cloud9.io.fastutil.Int2IntOpenHashMapWritableTest
  • edu.umd.cloud9.io.fastutil.String2FloatOpenHashMapWritableTest
  • edu.umd.cloud9.io.fastutil.String2IntOpenHashMapWritableTest
  • edu.umd.cloud9.io.map.HMapIDW
  • edu.umd.cloud9.io.map.HMapIDWTest
    • 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.