Package org.terrier.structures

Examples of org.terrier.structures.DocumentIndexEntry

439
440
441
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
517
518
519
      String outputPrefix = "-0";
      DataInputStream currentStream = new DataInputStream(Files.openFileStream(index.getPath() + ApplicationSetup.FILE_SEPARATOR + index.getPrefix() + "." + targetStructureName +outputPrefix+ ".pointers"));
      //logger.info("Adding pointers to the document index");
      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);
        }
        currentStream.close();
        Files.delete(index.getPath() + ApplicationSetup.FILE_SEPARATOR + index.getPrefix() + "." + targetStructureName +outputPrefix+ ".pointers");
        //logger.info("Copying direct file part for reduce task " + reduce);
View Full Code Here
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
      //write the offsets to the DocumentIndex
      final DataInputStream dis = new DataInputStream(Files.openFileStream(offsetsFilename));
      final DocumentIndexBuilder dios = new DocumentIndexBuilder(index, "document-df");
      final Iterator<DocumentIndexEntry> docidInput = (Iterator<DocumentIndexEntry>)index.getIndexStructureInputStream("document");
     
      DocumentIndexEntry die = null;
      int docid = 0;
        while (docidInput.hasNext())
      {
          DocumentIndexEntry old = docidInput.next();
          if (fieldCount == 0)
          {
            die = new BasicDocumentIndexEntry(old);
          }
          else
View Full Code Here
112
113
114
115
116
117
118
119
120
121
    }

    /** Return a DocumentIndexEntry for this document */
    public DocumentIndexEntry getDocumentStatistics()
  {
    DocumentIndexEntry die = new BasicDocumentIndexEntry();
    die.setDocumentLength(this.getDocumentLength());
    die.setNumberOfEntries(this.getNumberOfPointers());
    return die;
  }
View Full Code Here
361
362
363
364
365
366
367
368
369
370
371
      numberOfDocsSinceCheck++;
      numberOfDocsSinceFlush++;
     
      checkFlush();
      mp.addTerms(termsInDocument, currentId);
      DocumentIndexEntry die = termsInDocument.getDocumentStatistics();
      docIndexBuilder.addEntryToBuffer((FieldScore.FIELDS_COUNT > 0) ? die : new SimpleDocumentIndexEntry(die));
      metaBuilder.writeDocumentEntry(docProperties);
      currentId++;
      numberOfDocuments++;
    }
View Full Code Here
377
378
379
380
381
382
383
384
385
386
387
    lexiconBuilder.addDocumentTerms(_termsInDocument);
    /* add doc postings to the direct index */
    BitIndexPointer dirIndexPost = directIndexBuilder.writePostings(_termsInDocument.getPostings2());
      //.addDocument(termsInDocument.getPostings());
    /* add doc to documentindex */
    DocumentIndexEntry die = _termsInDocument.getDocumentStatistics();
    die.setBitIndexPointer(dirIndexPost);
    docIndexBuilder.addEntryToBuffer(die);
    /** add doc metadata to index */
    metaBuilder.writeDocumentEntry(docProperties);   
  }
 
View Full Code Here
493
494
495
496
497
498
499
500
501
502
503
    /* add words to lexicontree */
    lexiconBuilder.addDocumentTerms(_termsInDocument);
    /* add doc postings to the direct index */
    BitIndexPointer dirIndexPost = directIndexBuilder.writePostings(_termsInDocument.getPostings2());
    /* add doc to documentindex */
    DocumentIndexEntry die = _termsInDocument.getDocumentStatistics();
    die.setBitIndexPointer(dirIndexPost);
    docIndexBuilder.addEntryToBuffer(die);
    /** add doc metadata to index */
    metaBuilder.writeDocumentEntry(docProperties);
  }
 
View Full Code Here
435
436
437
438
439
440
441
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
      int sourceDocid = 0;
      //traversing the direct index, without any change
      while(docidInput1.hasNext())
      {
        BitIndexPointer pointerDF = emptyPointer;
        DocumentIndexEntry die = docidInput1.next();
        if (die.getDocumentLength() > 0)
        {
          pointerDF = dfOutput.writePostings(dfInput1.next());
        }
        die.setBitIndexPointer(pointerDF);
        docidOutput.addEntryToBuffer(die);
        metaBuilder.writeDocumentEntry(metaInput1.getAllItems(sourceDocid));
        sourceDocid++;
      }
      dfInput1.close();
      metaInput1.close();
      IndexUtil.close(docidInput1);
      final Iterator<DocumentIndexEntry> docidInput2 = (Iterator<DocumentIndexEntry>)srcIndex2.getIndexStructureInputStream("document");
      final PostingIndexInputStream dfInput2 = (PostingIndexInputStream)srcIndex2.getIndexStructureInputStream("direct");
      final MetaIndex metaInput2 = srcIndex2.getMetaIndex();
     
      sourceDocid = 0;
      while (docidInput2.hasNext())
      {
        DocumentIndexEntry die = docidInput2.next();
     
        BitIndexPointer pointerDF = emptyPointer;
        if (die.getDocumentLength() > 0)
        {
          final IterablePosting postings = dfInput2.next();
         
          List<Posting> postingList = new ArrayList<Posting>();
          while(postings.next() != IterablePosting.EOL)
          {
            final Posting p = postings.asWritablePosting();
            p.setId(termcodeHashmap.get(postings.getId()));
            postingList.add(p);
          }
          Collections.sort(postingList, new PostingIdComparator());
          pointerDF = dfOutput.writePostings(postingList.iterator());
        }
        die.setBitIndexPointer(pointerDF);
        docidOutput.addEntryToBuffer(die);
        metaBuilder.writeDocumentEntry(metaInput2.getAllItems(sourceDocid));
        sourceDocid++;
      }
      dfInput2.close();
View Full Code Here
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
     
      //traversing the first set of files, without any change
      while(docidInput1.hasNext())
      {
        metaInput1.hasNext();
        DocumentIndexEntry die = docidInput1.next();
        DocumentIndexEntry dieNew = (fieldCount > 0) ? die : new SimpleDocumentIndexEntry(die);
        docidOutput.addEntryToBuffer(dieNew);
        metaBuilder.writeDocumentEntry(metaInput1.next());
      }
     
      final Iterator<DocumentIndexEntry> docidInput2 = (Iterator<DocumentIndexEntry>)srcIndex2.getIndexStructureInputStream("document");
      final Iterator<String[]> metaInput2 = (Iterator<String[]>)srcIndex2.getIndexStructureInputStream("meta");
      //traversing the 2nd set of files, without any change
      while(docidInput2.hasNext())
      {
        metaInput2.hasNext();
        DocumentIndexEntry die = docidInput2.next();
        DocumentIndexEntry dieNew = (fieldCount > 0) ? die : new SimpleDocumentIndexEntry(die);
        docidOutput.addEntryToBuffer(dieNew);
        metaBuilder.writeDocumentEntry(metaInput2.next());
      }
     
      docidOutput.finishedCollections();
View Full Code Here
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
       */
      DirectIndex directIndex = index.getDirectIndex();
      assertNotNull(directIndex);
      // for each document
      for (int d = 0; d < dirTfs.length; d++) {
        DocumentIndexEntry de = documentIndex.getDocumentEntry(d);
        assertNotNull(de);
        ip = directIndex.getPostings((BitIndexPointer) de);
        FieldPosting fp = fieldsExpected ? (FieldPosting)ip : null;
        // for each term
        int t = 0;
        int countFoundTerms = 0;
        while (ip.next() != IterablePosting.EOL) {
          int termid = ip.getId();
          assertTrue(termid >= 0);
          String term = lexicon.getLexiconEntry(termid).getKey();
          assertNotNull(term);
          countFoundTerms++;
          assertTrue(dirTfs[d].containsKey(term));
          assertEquals(dirTfs[d].get(term), ip.getFrequency());
          assertEquals(doclens[d], ip.getDocumentLength());         
         
          if (fieldsExpected) {
            assertEquals(2, fp.getFieldFrequencies().length);
            for (int f = 0; f < 2; f++) {
              assertEquals(dirFfs[d].get(term)[f], fp.getFieldFrequencies()[f]);
            }
          }
          t++;
        }
        assertEquals(dirTfs[d].size() ,countFoundTerms);
        ip.close();
      }
      // post-check
      assertEquals(IterablePosting.EOL, ip.next());

      /**
       * Test {@link IterablePosting} entries from a {@link DirectIndexInputStream}
       */
      bpiis = (BitPostingIndexInputStream) index.getIndexStructureInputStream("direct");
      assertNotNull(bpiis);
      // for each document
      for (int d = 0; d < dirTfs.length; d++) {
        assertTrue(bpiis.hasNext());
        ip = bpiis.next();
        assertNotNull(ip);
        FieldPosting fp = fieldsExpected ? (FieldPosting)ip : null;
        // for each term
        int t = 0;
        int countFoundTerms = 0;
        while (ip.next() != IterablePosting.EOL) {
          int termid = ip.getId();
          assertTrue(termid >= 0);
          String term = lexicon.getLexiconEntry(termid).getKey();
          assertNotNull(term);
          countFoundTerms++;
          assertTrue(dirTfs[d].containsKey(term));
          assertEquals(dirTfs[d].get(term), ip.getFrequency());
          assertEquals(doclens[d], ip.getDocumentLength());         
         
          if (fieldsExpected) {
            assertEquals(2, fp.getFieldFrequencies().length);
            for (int f = 0; f < 2; f++) {
              assertEquals(dirFfs[d].get(term)[f], fp.getFieldFrequencies()[f]);
            }
          }
          t++;
        }
        assertEquals(dirTfs[d].size() ,countFoundTerms);
      }
      // post-check
      assertFalse(bpiis.hasNext());

      /**
       * Test posting array entries from a {@link DirectIndex}
       */
      // for each document
      for (int d = 0; d < dirTfs.length; d++) {
        DocumentIndexEntry de = documentIndex.getDocumentEntry(d);
        assertNotNull(de);
       
        int[][] terms = directIndex.getTerms(de);
       
        if (!fieldsExpected) {
View Full Code Here
354
355
356
357
358
359
360
361
362
363
364
365
366
367
    //check index as stream
    Iterator<DocumentIndexEntry> iDie = (Iterator<DocumentIndexEntry>) index.getIndexStructureInputStream("document");
    assertNotNull("Failed to get a document inputstream", iDie);
    while(iDie.hasNext())
    {
      DocumentIndexEntry die = iDie.next();
      docid++;
      //System.out.println(die.getDocumentLength());
      assertEquals("Document lengths for docid "+ docid + " dont match", lengths[docid], die.getDocumentLength());
    }
    //check docid is as large as expected
    assertEquals("Metaindex as stream didnt have expected number of entries", lengths.length -1, docid);
    IndexUtil.close(iDie);
   
View Full Code Here
TOP

Related Classes of org.terrier.structures.DocumentIndexEntry

  • org.terrier.indexing.BasicIndexer
  • org.terrier.indexing.BasicSinglePassIndexer
  • org.terrier.indexing.BlockIndexer
  • org.terrier.indexing.TestIndexers
  • org.terrier.structures.indexing.DocumentPostingList
  • org.terrier.structures.indexing.singlepass.hadoop.Inv2DirectMultiReduce
  • org.terrier.structures.indexing.singlepass.hadoop.TestBitPostingIndexInputFormat
  • org.terrier.structures.indexing.singlepass.Inverted2DirectIndexBuilder
  • org.terrier.structures.merging.StructureMerger
  • org.terrier.tests.ShakespeareEndToEndTest
    • 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.