Examples of DocsAndPositionsEnum

org.apache.lucene.index.DocsAndPositionsEnum
Also iterates through positions.

Examples of org.apache.lucene.index.DocsAndPositionsEnum

    MemoryIndex memory = new MemoryIndex(true,  random().nextInt(50) * 1024 * 1024);
    for (int i = 0; i < numIters; i++) { // check reuse
      memory.addField("foo", "bar", analyzer);
      AtomicReader reader = (AtomicReader) memory.createSearcher().getIndexReader();
      assertEquals(1, reader.terms("foo").getSumTotalTermFreq());
      DocsAndPositionsEnum disi = reader.termPositionsEnum(new Term("foo", "bar"));
      int docid = disi.docID();
      assertEquals(-1, docid);
      assertTrue(disi.nextDoc() != DocIdSetIterator.NO_MORE_DOCS);
      assertEquals(0, disi.nextPosition());
      assertEquals(0, disi.startOffset());
      assertEquals(3, disi.endOffset());
      
      // now reuse and check again
      TermsEnum te = reader.terms("foo").iterator(null);
      assertTrue(te.seekExact(new BytesRef("bar")));
      disi = te.docsAndPositions(null, disi);
      docid = disi.docID();
      assertEquals(-1, docid);
      assertTrue(disi.nextDoc() != DocIdSetIterator.NO_MORE_DOCS);
      reader.close();
      memory.reset();
    }
  }

View Full Code Here

Examples of org.apache.lucene.index.DocsAndPositionsEnum

      if (termState == null) {
        // Term doesn't exist in reader
        continue;
      }
      termsEnum.seekExact(term.bytes(), termState);
      DocsAndPositionsEnum postings = termsEnum.docsAndPositions(liveDocs, null, DocsEnum.FLAG_NONE);
      if (postings == null) {
        // term does exist, but has no positions
        throw new IllegalStateException("field \"" + term.field() + "\" was indexed without position data; cannot run PhraseQuery (term=" + term.text() + ")");
      }
      cost += postings.cost();
      docsEnums.add(postings);
    }


    _queue = new DocsQueue(docsEnums);
    _posList = new IntQueue();

View Full Code Here

Examples of org.apache.lucene.index.DocsAndPositionsEnum

    // waste CPU merging them:
    _posList.clear();
    _doc = _queue.top().docID();


    // merge sort all positions together
    DocsAndPositionsEnum postings;
    do {
      postings = _queue.top();


      final int freq = postings.freq();
      for (int i = 0; i < freq; i++) {
        _posList.add(postings.nextPosition());
      }


      if (postings.nextDoc() != NO_MORE_DOCS) {
        _queue.updateTop();
      } else {
        _queue.pop();
      }
    } while (_queue.size() > 0 && _queue.top().docID() == _doc);

View Full Code Here

Examples of org.apache.lucene.index.DocsAndPositionsEnum

  }


  @Override
  public final int advance(int target) throws IOException {
    while (_queue.top() != null && target > _queue.top().docID()) {
      DocsAndPositionsEnum postings = _queue.pop();
      if (postings.advance(target) != NO_MORE_DOCS) {
        _queue.add(postings);
      }
    }
    return nextDoc();
  }

View Full Code Here

Examples of org.apache.lucene.index.DocsAndPositionsEnum

    DocsQueue(List<DocsAndPositionsEnum> docsEnums) throws IOException {
      super(docsEnums.size());


      Iterator<DocsAndPositionsEnum> i = docsEnums.iterator();
      while (i.hasNext()) {
        DocsAndPositionsEnum postings = i.next();
        if (postings.nextDoc() != DocIdSetIterator.NO_MORE_DOCS) {
          add(postings);
        }
      }
    }

View Full Code Here

Examples of org.apache.lucene.index.DocsAndPositionsEnum

        setOther(postings, reuse); // postings.other = reuse 
      }
      return postings.reset(liveDocs, termState);
    } else {
      if (reuse instanceof PulsingDocsAndPositionsEnum) {
        DocsAndPositionsEnum wrapped = wrappedPostingsReader.docsAndPositions(field, termState.wrappedTermState, liveDocs, (DocsAndPositionsEnum) getOther(reuse),
                                                                              flags);
        setOther(wrapped, reuse); // wrapped.other = reuse
        return wrapped;
      } else {
        return wrappedPostingsReader.docsAndPositions(field, termState.wrappedTermState, liveDocs, reuse, flags);

View Full Code Here

Examples of org.apache.lucene.index.DocsAndPositionsEnum

      final TermsEnum termsEnum = fieldTerms.iterator(null);


      for (int pos=0; pos<postingsFreqs.length; pos++) {
        Term[] terms = termArrays.get(pos);


        final DocsAndPositionsEnum postingsEnum;
        int docFreq;


        if (terms.length > 1) {
          postingsEnum = new UnionDocsAndPositionsEnum(liveDocs, context, terms, termContexts, termsEnum);

View Full Code Here

Examples of org.apache.lucene.index.DocsAndPositionsEnum


      String docTerm = term.utf8ToString();


      if (matchedTokens.contains(docTerm) || docTerm.startsWith(prefixToken)) {


        DocsAndPositionsEnum docPosEnum = it.docsAndPositions(null, null, DocsAndPositionsEnum.FLAG_OFFSETS);
        docPosEnum.nextDoc();


        // use the first occurrence of the term
        int p = docPosEnum.nextPosition();
        if (p < position) {
          position = p;
        }
      }
    }

View Full Code Here

Examples of org.apache.lucene.index.DocsAndPositionsEnum

        if (state == null) { /* term doesnt exist in this segment */
          assert termNotInReader(reader, t): "no termstate found but term exists in reader";
          return null;
        }
        te.seekExact(t.bytes(), state);
        DocsAndPositionsEnum postingsEnum = te.docsAndPositions(liveDocs, null, DocsEnum.FLAG_NONE);


        // PhraseQuery on a field that did not index
        // positions.
        if (postingsEnum == null) {
          assert te.seekExact(t.bytes()) : "termstate found but no term exists in reader";

View Full Code Here

Examples of org.apache.lucene.index.DocsAndPositionsEnum

    
  private Map<Integer,Object> highlightField(String field, String contents[], BreakIterator bi, BytesRef terms[], int[] docids, List<AtomicReaderContext> leaves, int maxPassages, Query query) throws IOException {  
    Map<Integer,Object> highlights = new HashMap<Integer,Object>();
    
    // reuse in the real sense... for docs in same segment we just advance our old enum
    DocsAndPositionsEnum postings[] = null;
    TermsEnum termsEnum = null;
    int lastLeaf = -1;


    PassageFormatter fieldFormatter = getFormatter(field);
    if (fieldFormatter == null) {
      throw new NullPointerException("PassageFormatter cannot be null");
    }
    
    // check if we should do any multitermprocessing
    Analyzer analyzer = getIndexAnalyzer(field);
    CharacterRunAutomaton automata[] = new CharacterRunAutomaton[0];
    if (analyzer != null) {
      automata = MultiTermHighlighting.extractAutomata(query, field);
    }
    
    final BytesRef allTerms[];
    if (automata.length > 0) {
      allTerms = new BytesRef[terms.length + 1];
      System.arraycopy(terms, 0, allTerms, 0, terms.length);
    } else {
      allTerms = terms;
    }


    for (int i = 0; i < docids.length; i++) {
      String content = contents[i];
      if (content.length() == 0) {
        continue; // nothing to do
      }
      bi.setText(content);
      int doc = docids[i];
      int leaf = ReaderUtil.subIndex(doc, leaves);
      AtomicReaderContext subContext = leaves.get(leaf);
      AtomicReader r = subContext.reader();
      Terms t = r.terms(field);
      if (t == null) {
        continue; // nothing to do
      }
      if (leaf != lastLeaf) {
        termsEnum = t.iterator(null);
        postings = new DocsAndPositionsEnum[allTerms.length];
      }
      if (automata.length > 0) {
        DocsAndPositionsEnum dp = MultiTermHighlighting.getDocsEnum(analyzer.tokenStream(field, content), automata);
        dp.advance(doc - subContext.docBase);
        postings[terms.length] = dp;
      }
      Passage passages[] = highlightDoc(field, allTerms, content.length(), bi, doc - subContext.docBase, termsEnum, postings, maxPassages);
      if (passages.length == 0) {
        passages = getEmptyHighlight(field, bi, maxPassages);

View Full Code Here

0 1 2 3 4 5

TOP

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.