Package org.terrier.structures

Examples of org.terrier.structures.FieldDocumentIndexEntry

442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
      while(diis.hasNext())
      {
        DocumentIndexEntry die =  diis.next();
        pointer.readFields(currentStream);
        DocumentIndexEntry newDIentry = fields
          ? new FieldDocumentIndexEntry(die)
          : new BasicDocumentIndexEntry(die);
        newDIentry.setOffset(pointer);
        newDIentry.setNumberOfEntries(pointer.getNumberOfEntries());
        dios.addEntryToBuffer(newDIentry);
      }
      //logger.info("Renaming reducer output as direct file");
      Files.delete(index.getPath() + ApplicationSetup.FILE_SEPARATOR + index.getPrefix() + "." + targetStructureName+ BitIn.USUAL_EXTENSION);
      Files.rename(
          index.getPath() + ApplicationSetup.FILE_SEPARATOR + index.getPrefix() + "." + targetStructureName+outputPrefix + BitIn.USUAL_EXTENSION,
          index.getPath() + ApplicationSetup.FILE_SEPARATOR + index.getPrefix() + "." + targetStructureName+ BitIn.USUAL_EXTENSION);
      currentStream.close();
      Files.delete(index.getPath() + ApplicationSetup.FILE_SEPARATOR + index.getPrefix() + "." + targetStructureName +outputPrefix+ ".pointers");
    }
    else if (numberOfReducers <= numberOfReduceTaskLimits)
    {
      //logger.info("Merging direct index pointers from "+ numberOfReducers + " reducers");
      final int partitionSize = (int)Math.ceil( (double)(index.getCollectionStatistics().getNumberOfDocuments()) / (double)numberOfReducers);     
      for(byte reduce = 0; reduce < numberOfReducers; reduce++)
      {
        //logger.info("Merging in pointers from reduce task " + reduce);
        String outputPrefix = "-" + reduce;
        DataInputStream currentStream = new DataInputStream(Files.openFileStream(index.getPath() + ApplicationSetup.FILE_SEPARATOR + index.getPrefix() + "." + targetStructureName +outputPrefix+ ".pointers"));
        for(int docOffset = 0; docOffset < partitionSize && diis.hasNext(); docOffset++)
        {
          DocumentIndexEntry die =  diis.next();
          pointer.readFields(currentStream);
          DocumentIndexEntry newDIentry = fields
            ? new FieldDocumentIndexEntry(die)
            : new BasicDocumentIndexEntry(die);
          newDIentry.setOffset(pointer);
          newDIentry.setFileNumber(reduce);
          newDIentry.setNumberOfEntries(pointer.getNumberOfEntries());
          dios.addEntryToBuffer(newDIentry);
        }
        currentStream.close();
        Files.delete(index.getPath() + ApplicationSetup.FILE_SEPARATOR + index.getPrefix() + "." + targetStructureName +outputPrefix+ ".pointers");
        //logger.info("Renaming direct file part for reduce task " + reduce);
        String sourcePartDFfilename = index.getPath() + ApplicationSetup.FILE_SEPARATOR + index.getPrefix() + "." + targetStructureName+outputPrefix + BitIn.USUAL_EXTENSION;
        String destPartDFfilename = index.getPath() + ApplicationSetup.FILE_SEPARATOR + index.getPrefix() + "." + targetStructureName+ BitIn.USUAL_EXTENSION + reduce;       
        Files.rename(sourcePartDFfilename, destPartDFfilename);
      }
      index.setIndexProperty("index."+targetStructureName+".data-files", ""+numberOfReducers);
      index.flush();
      IndexUtil.close(diis);
    }
    else
    {
      //logger.info("Merging direct index output from "+ numberOfReducers + " reducers");
     
      final int partitionSize = (int)Math.ceil( (double)(index.getCollectionStatistics().getNumberOfDocuments()) / (double)numberOfReducers);
      final OutputStream DFout = Files.writeFileStream(index.getPath() + ApplicationSetup.FILE_SEPARATOR + index.getPrefix() + "." + targetStructureName+ BitIn.USUAL_EXTENSION);
      long finalFileOffset = 0;
     
      for(int reduce = 0; reduce < numberOfReducers; reduce++)
      {
        //logger.info("Copying document index part for reduce task " + reduce);
        String outputPrefix = "-" + reduce;
        DataInputStream currentStream = new DataInputStream(Files.openFileStream(index.getPath() + ApplicationSetup.FILE_SEPARATOR + index.getPrefix() + "." + targetStructureName +outputPrefix+ ".pointers"));
        for(int docOffset = 0; docOffset < partitionSize && diis.hasNext(); docOffset++)
        {
          DocumentIndexEntry die =  diis.next();
          pointer.readFields(currentStream);
          DocumentIndexEntry newDIentry = fields
            ? new FieldDocumentIndexEntry(die)
            : new BasicDocumentIndexEntry(die);
          newDIentry.setOffset(finalFileOffset + pointer.getOffset(), pointer.getOffsetBits());
          newDIentry.setNumberOfEntries(pointer.getNumberOfEntries());
          dios.addEntryToBuffer(newDIentry);
        }
View Full Code Here
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
        return new int[0];
      }
    }
    else
    {
      FieldDocumentIndexEntry fdie = null;
      try{
        fdie = ((FieldDocumentIndexEntry)doi.getDocumentEntry(id));
      } catch (IOException ioe) {
        //TODO log?
        System.err.println("Problem looking for doclength for document "+ id);
        ioe.printStackTrace();
        return new int[0];
      }
      return fdie.getFieldLengths();
    }
  }
View Full Code Here
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
        return new int[0];
      }
    }
    else
    {
      FieldDocumentIndexEntry fdie = null;
      try{
        fdie = ((FieldDocumentIndexEntry)doi.getDocumentEntry(id));
      } catch (IOException ioe) {
        //TODO log?
        System.err.println("Problem looking for doclength for document "+ id);
        ioe.printStackTrace();
        return new int[0];
      }
      return fdie.getFieldLengths();
    }
  }
View Full Code Here
129
130
131
132
133
134
135
136
137
138
139
  /**
   * {@inheritDoc}
   */
  public DocumentIndexEntry getDocumentStatistics()
  {
    FieldDocumentIndexEntry fdie = new FieldDocumentIndexEntry(this.fieldCount);
    fdie.setDocumentLength(documentLength);
    fdie.setNumberOfEntries(occurrences.size());
    fdie.setFieldLengths(fieldLengths);
    return fdie;
  }
View Full Code Here
291
292
293
294
295
296
297
    start = true;
    createMemoryPostings();
    super.emptyDocIndexEntry = new SimpleDocumentIndexEntry();
    super.docIndexBuilder = new DocumentIndexBuilder(currentIndex, "document");
    super.metaBuilder = createMetaIndexBuilder();
    emptyDocIndexEntry = (FieldScore.FIELDS_COUNT > 0) ? new FieldDocumentIndexEntry(FieldScore.FIELDS_COUNT) : new SimpleDocumentIndexEntry();
  }
View Full Code Here
176
177
178
179
180
181
182
183
184
185
186
    createMemoryPostings();
    currentIndex = Index.createNewIndex(path, prefix);
    docIndexBuilder = new DocumentIndexBuilder(currentIndex, "document");
    metaBuilder = createMetaIndexBuilder();
   
    emptyDocIndexEntry = (FieldScore.FIELDS_COUNT > 0) ? new FieldDocumentIndexEntry(FieldScore.FIELDS_COUNT) : new SimpleDocumentIndexEntry();
   
    MAX_DOCS_PER_BUILDER = Integer.parseInt(ApplicationSetup.getProperty("indexing.max.docs.per.builder", "0"));
    maxMemory = Long.parseLong(ApplicationSetup.getProperty("indexing.singlepass.max.postings.memory", "0"));
    final boolean boundaryDocsEnabled = BUILDER_BOUNDARY_DOCUMENTS.size() > 0;
    final int collections_length = collections.length;
View Full Code Here
207
208
209
210
211
212
213
214
215
216
217
  //    logger.error("Cannot make DirectInvertedOutputStream:", ioe);
    }
      //  new DirectIndexBuilder(currentIndex, "direct");
    docIndexBuilder = new DocumentIndexBuilder(currentIndex, "document");
    metaBuilder = createMetaIndexBuilder();
    emptyDocIndexEntry = (FieldScore.FIELDS_COUNT > 0) ? new FieldDocumentIndexEntry(FieldScore.FIELDS_COUNT) : new BasicDocumentIndexEntry();
       
    //int LexiconCount = 0;
    int numberOfDocuments = 0; int numberOfTokens = 0;
    //final long startBunchOfDocuments = System.currentTimeMillis();
    final int collections_length = collections.length;
View Full Code Here
333
334
335
336
337
338
339
340
341
342
343
    } catch (IOException ioe) {
  //    logger.error("Cannot make DirectInvertedOutputStream:", ioe);
    }
    docIndexBuilder = new DocumentIndexBuilder(currentIndex, "document");
    metaBuilder = createMetaIndexBuilder();
    emptyDocIndexEntry = (FieldScore.FIELDS_COUNT > 0) ? new FieldDocumentIndexEntry(FieldScore.FIELDS_COUNT) : new BasicDocumentIndexEntry();
   
    //int LexiconCount = 0;
    int numberOfDocuments = 0;
    int numberOfTokens = 0;
    //long startBunchOfDocuments = System.currentTimeMillis();
View Full Code Here
109
110
111
112
113
114
115
116
117
118
119
    createMemoryPostings();
    currentIndex = Index.createNewIndex(path, prefix);
    docIndexBuilder = new DocumentIndexBuilder(currentIndex, "document");
    metaBuilder = createMetaIndexBuilder();
   
    emptyDocIndexEntry = (FieldScore.FIELDS_COUNT > 0) ? new FieldDocumentIndexEntry(FieldScore.FIELDS_COUNT) : new SimpleDocumentIndexEntry();
   
    MAX_DOCS_PER_BUILDER = Integer.parseInt(ApplicationSetup.getProperty("indexing.max.docs.per.builder", "0"));
    maxMemory = Long.parseLong(ApplicationSetup.getProperty("indexing.singlepass.max.postings.memory", "0"));
    final boolean boundaryDocsEnabled = BUILDER_BOUNDARY_DOCUMENTS.size() > 0;
    final int collections_length = collections.length;
View Full Code Here
TOP

Related Classes of org.terrier.structures.FieldDocumentIndexEntry

  • org.terrier.indexing.BasicIndexer
  • org.terrier.indexing.BasicSinglePassIndexer
  • org.terrier.indexing.BlockIndexer
  • org.terrier.indexing.ExtensibleSinglePassIndexer
  • org.terrier.indexing.hadoop.Hadoop_BasicSinglePassIndexer
  • org.terrier.structures.indexing.FieldDocumentPostingList
  • org.terrier.structures.indexing.singlepass.hadoop.Inv2DirectMultiReduce
  • org.terrier.structures.postings.BlockFieldIterablePosting
  • org.terrier.structures.postings.FieldIterablePosting
    • 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.