Examples of org.apache.lucene.index.DocsEnum

org.apache.lucene.index.DocsEnum
Iterates through the documents and term freqs. NOTE: you must first call {@link #nextDoc} before usingany of the per-doc methods.


  @Test
  public void testTermDocIterable() throws IOException {
    for (int pass = 0; pass < 1; pass++) {
      for (int id = 0; id < BLOCKS; id++) {
        DocsEnum termDocs = reader.termDocsEnum(new Term("id", Integer.toString(id)));
        TermDocIterable iterable = new TermDocIterable(termDocs, reader);
        int count = 0;
        int i = 0;
        long s = System.nanoTime();
        for (Document document : iterable) {

View Full Code Here

                return getResult();
              }
            }
          } while (nextDoc != DocsEnum.NO_MORE_DOCS);
        } else {
          DocsEnum docsEnum = termsEnum.docs(_liveDocs, null);
          int nextDoc;
          do {
            nextDoc = docsEnum.nextDoc();
            if (traceComplete()) {
              return getResult();
            }
          } while (nextDoc != DocsEnum.NO_MORE_DOCS);
        }

View Full Code Here

          }
        });
        LOG.debug("Prime Doc BitSet missing for segment [" + reader + "] current size [" + primeDocMap.size() + "]");
        final OpenBitSet bs = new OpenBitSet(reader.maxDoc());


        DocsEnum termDocsEnum = reader.termDocsEnum(primeDocTerm);
        if (termDocsEnum == null) {
          return bs;
        }
        int docFreq = reader.docFreq(primeDocTerm);
        int doc;
        int count = 0;
        while ((doc = termDocsEnum.nextDoc()) != DocIdSetIterator.NO_MORE_DOCS) {
          bs.fastSet(doc);
          count++;
        }
        if (count == docFreq) {
          primeDocMap.put(key, bs);

View Full Code Here

    Terms terms = fields.terms(BlurConstants.FAMILY);
    TermsEnum iterator = terms.iterator(null);
    BytesRef text = new BytesRef(family);
    int lastDocId = primeDocRowId + numberOfDocsInRow;
    if (iterator.seekExact(text, true)) {
      DocsEnum docs = iterator.docs(liveDocs, null, DocsEnum.FLAG_NONE);
      int doc = primeDocRowId;
      while ((doc = docs.advance(doc)) < lastDocId) {
        bits.set(doc - primeDocRowId);
      }
    }
  }

View Full Code Here

    }


    // If we're still here, we have a cache miss. We need to fetch the
    // value from disk, and then also put it in the cache:
    int ret = TaxonomyReader.INVALID_ORDINAL;
    DocsEnum docs = MultiFields.getTermDocsEnum(indexReader, null, Consts.FULL, new BytesRef(cp.toString(delimiter)), 0);
    if (docs != null && docs.nextDoc() != DocIdSetIterator.NO_MORE_DOCS) {
      ret = docs.docID();
      
      // we only store the fact that a category exists, not its inexistence.
      // This is required because the caches are shared with new DTR instances
      // that are allocated from doOpenIfChanged. Therefore, if we only store
      // information about found categories, we cannot accidently tell a new

View Full Code Here

    int doc = -1;
    DirectoryReader reader = readerManager.acquire();
    try {
      final BytesRef catTerm = new BytesRef(categoryPath.toString(delimiter));
      TermsEnum termsEnum = null; // reuse
      DocsEnum docs = null; // reuse
      for (AtomicReaderContext ctx : reader.leaves()) {
        Terms terms = ctx.reader().terms(Consts.FULL);
        if (terms != null) {
          termsEnum = terms.iterator(termsEnum);
          if (termsEnum.seekExact(catTerm, true)) {
            // liveDocs=null because the taxonomy has no deletes
            docs = termsEnum.docs(null, docs, 0 /* freqs not required */);
            // if the term was found, we know it has exactly one document.
            doc = docs.nextDoc() + ctx.docBase;
            break;
          }
        }
      }
    } finally {

View Full Code Here


    boolean aborted = false;
    DirectoryReader reader = readerManager.acquire();
    try {
      TermsEnum termsEnum = null;
      DocsEnum docsEnum = null;
      for (AtomicReaderContext ctx : reader.leaves()) {
        Terms terms = ctx.reader().terms(Consts.FULL);
        if (terms != null) { // cannot really happen, but be on the safe side
          termsEnum = terms.iterator(termsEnum);
          while (termsEnum.next() != null) {
            if (!cache.isFull()) {
              BytesRef t = termsEnum.term();
              // Since we guarantee uniqueness of categories, each term has exactly
              // one document. Also, since we do not allow removing categories (and
              // hence documents), there are no deletions in the index. Therefore, it
              // is sufficient to call next(), and then doc(), exactly once with no
              // 'validation' checks.
              CategoryPath cp = new CategoryPath(t.utf8ToString(), delimiter);
              docsEnum = termsEnum.docs(null, docsEnum, DocsEnum.FLAG_NONE);
              boolean res = cache.put(cp, docsEnum.nextDoc() + ctx.docBase);
              assert !res : "entries should not have been evicted from the cache";
            } else {
              // the cache is full and the next put() will evict entries from it, therefore abort the iteration.
              aborted = true;
              break;

View Full Code Here

      final int size = r.numDocs();
      final OrdinalMap ordinalMap = map;
      ordinalMap.setSize(size);
      int base = 0;
      TermsEnum te = null;
      DocsEnum docs = null;
      for (final AtomicReaderContext ctx : r.leaves()) {
        final AtomicReader ar = ctx.reader();
        final Terms terms = ar.terms(Consts.FULL);
        te = terms.iterator(te);
        while (te.next() != null) {
          String value = te.term().utf8ToString();
          CategoryPath cp = new CategoryPath(value, delimiter);
          final int ordinal = addCategory(cp);
          docs = te.docs(null, docs, DocsEnum.FLAG_NONE);
          ordinalMap.addMapping(docs.nextDoc() + base, ordinal);
        }
        base += ar.maxDoc(); // no deletions, so we're ok
      }
      ordinalMap.addDone();
    } finally {

View Full Code Here

      this.matchingDocsIterator = matchingDocs.iterator();
    }


    protected void fillDocsAndScores(FixedBitSet matchingDocs, Bits acceptDocs, TermsEnum termsEnum) throws IOException {
      BytesRef spare = new BytesRef();
      DocsEnum docsEnum = null;
      for (int i = 0; i < terms.size(); i++) {
        if (termsEnum.seekExact(terms.get(ords[i], spare), true)) {
          docsEnum = termsEnum.docs(acceptDocs, docsEnum, DocsEnum.FLAG_NONE);
          float score = TermsIncludingScoreQuery.this.scores[ords[i]];
          for (int doc = docsEnum.nextDoc(); doc != DocIdSetIterator.NO_MORE_DOCS; doc = docsEnum.nextDoc()) {
            matchingDocs.set(doc);
            // In the case the same doc is also related to a another doc, a score might be overwritten. I think this
            // can only happen in a many-to-many relation
            scores[doc] = score;
          }

View Full Code Here

    }


    @Override
    protected void fillDocsAndScores(FixedBitSet matchingDocs, Bits acceptDocs, TermsEnum termsEnum) throws IOException {
      BytesRef spare = new BytesRef();
      DocsEnum docsEnum = null;
      for (int i = 0; i < terms.size(); i++) {
        if (termsEnum.seekExact(terms.get(ords[i], spare), true)) {
          docsEnum = termsEnum.docs(acceptDocs, docsEnum, DocsEnum.FLAG_NONE);
          float score = TermsIncludingScoreQuery.this.scores[ords[i]];
          for (int doc = docsEnum.nextDoc(); doc != DocIdSetIterator.NO_MORE_DOCS; doc = docsEnum.nextDoc()) {
            // I prefer this:
            /*if (scores[doc] < score) {
              scores[doc] = score;
              matchingDocs.set(doc);
            }*/

View Full Code Here

0 1 2 3 4 5 6 7 8 9

TOP

Related Classes of org.apache.lucene.index.DocsEnum

com.basho.yokozuna.handler.EntropyData

com.browseengine.bobo.facets.data.FacetDataCache

com.browseengine.bobo.facets.data.MultiValueFacetDataCache

com.browseengine.bobo.facets.data.MultiValueWithWeightFacetDataCache

com.browseengine.bobo.facets.filter.AdaptiveFacetFilter$TermListRandomAccessDocIdSet$TermDocIdSet

com.browseengine.bobo.facets.impl.CompactMultiValueFacetHandler

com.pearson.entech.elasticsearch.search.facet.approx.termlist.TermListFacetExecutor$PostExecutor

org.apache.blur.lucene.search.PrimeDocCache

org.apache.blur.lucene.warmup.IndexTracer

org.apache.blur.utils.BlurUtil

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.