Examples with DocIdSet org.apache.lucene.search.DocIdSet used on opensource projects

Example 21 with DocIdSet

use of org.apache.lucene.search.DocIdSet in project lucene-solr by apache.

the class Filter method createWeight.

//
// Query compatibility
//
@Override
public Weight createWeight(IndexSearcher searcher, boolean needsScores, float boost) throws IOException {
    return new Weight(this) {

        @Override
        public void extractTerms(Set<Term> terms) {
        }

        @Override
        public Explanation explain(LeafReaderContext context, int doc) throws IOException {
            final Scorer scorer = scorer(context);
            final boolean match = (scorer != null && scorer.iterator().advance(doc) == doc);
            if (match) {
                assert scorer.score() == 0f;
                return Explanation.match(0f, "Match on id " + doc);
            } else {
                return Explanation.match(0f, "No match on id " + doc);
            }
        }

        @Override
        public Scorer scorer(LeafReaderContext context) throws IOException {
            final DocIdSet set = getDocIdSet(context, null);
            if (set == null) {
                return null;
            }
            if (applyLazily && set.bits() != null) {
                final Bits bits = set.bits();
                final DocIdSetIterator approximation = DocIdSetIterator.all(context.reader().maxDoc());
                final TwoPhaseIterator twoPhase = new TwoPhaseIterator(approximation) {

                    @Override
                    public boolean matches() throws IOException {
                        return bits.get(approximation.docID());
                    }

                    @Override
                    public float matchCost() {
                        // TODO use cost of bits.get()
                        return 10;
                    }
                };
                return new ConstantScoreScorer(this, 0f, twoPhase);
            }
            final DocIdSetIterator iterator = set.iterator();
            if (iterator == null) {
                return null;
            }
            return new ConstantScoreScorer(this, 0f, iterator);
        }
    };
}

Example 22 with DocIdSet

use of org.apache.lucene.search.DocIdSet in project lucene-solr by apache.

the class QueryWrapperFilter method getDocIdSet.

@Override
public DocIdSet getDocIdSet(final LeafReaderContext context, final Bits acceptDocs) throws IOException {
    // get a private context that is used to rewrite, createWeight and score eventually
    final LeafReaderContext privateContext = context.reader().getContext();
    final Weight weight = new IndexSearcher(privateContext).createNormalizedWeight(query, false);
    DocIdSet set = new DocIdSet() {

        @Override
        public DocIdSetIterator iterator() throws IOException {
            Scorer scorer = weight.scorer(privateContext);
            return scorer == null ? null : scorer.iterator();
        }

        @Override
        public long ramBytesUsed() {
            return 0L;
        }
    };
    return BitsFilteredDocIdSet.wrap(set, acceptDocs);
}

Example 23 with DocIdSet

use of org.apache.lucene.search.DocIdSet in project lucene-solr by apache.

the class TestDocIdSetBuilder method testDense.

public void testDense() throws IOException {
    final int maxDoc = 1000000 + random().nextInt(1000000);
    DocIdSetBuilder builder = new DocIdSetBuilder(maxDoc);
    final int numIterators = 1 + random().nextInt(10);
    final FixedBitSet ref = new FixedBitSet(maxDoc);
    for (int i = 0; i < numIterators; ++i) {
        RoaringDocIdSet.Builder b = new RoaringDocIdSet.Builder(maxDoc);
        for (int doc = random().nextInt(1000); doc < maxDoc; doc += 1 + random().nextInt(100)) {
            b.add(doc);
            ref.set(doc);
        }
        builder.add(b.build().iterator());
    }
    DocIdSet result = builder.build();
    assertTrue(result instanceof BitDocIdSet);
    assertEquals(new BitDocIdSet(ref), result);
}

Example 24 with DocIdSet

use of org.apache.lucene.search.DocIdSet in project lucene-solr by apache.

the class TestDocIdSetBuilder method testLeverageStats.

public void testLeverageStats() throws IOException {
    // single-valued points
    PointValues values = new DummyPointValues(42, 42);
    DocIdSetBuilder builder = new DocIdSetBuilder(100, values, "foo");
    assertEquals(1d, builder.numValuesPerDoc, 0d);
    assertFalse(builder.multivalued);
    DocIdSetBuilder.BulkAdder adder = builder.grow(2);
    adder.add(5);
    adder.add(7);
    DocIdSet set = builder.build();
    assertTrue(set instanceof BitDocIdSet);
    assertEquals(2, set.iterator().cost());
    // multi-valued points
    values = new DummyPointValues(42, 63);
    builder = new DocIdSetBuilder(100, values, "foo");
    assertEquals(1.5, builder.numValuesPerDoc, 0d);
    assertTrue(builder.multivalued);
    adder = builder.grow(2);
    adder.add(5);
    adder.add(7);
    set = builder.build();
    assertTrue(set instanceof BitDocIdSet);
    // it thinks the same doc was added twice
    assertEquals(1, set.iterator().cost());
    // incomplete stats
    values = new DummyPointValues(42, -1);
    builder = new DocIdSetBuilder(100, values, "foo");
    assertEquals(1d, builder.numValuesPerDoc, 0d);
    assertTrue(builder.multivalued);
    values = new DummyPointValues(-1, 84);
    builder = new DocIdSetBuilder(100, values, "foo");
    assertEquals(1d, builder.numValuesPerDoc, 0d);
    assertTrue(builder.multivalued);
    // single-valued terms
    Terms terms = new DummyTerms(42, 42);
    builder = new DocIdSetBuilder(100, terms);
    assertEquals(1d, builder.numValuesPerDoc, 0d);
    assertFalse(builder.multivalued);
    adder = builder.grow(2);
    adder.add(5);
    adder.add(7);
    set = builder.build();
    assertTrue(set instanceof BitDocIdSet);
    assertEquals(2, set.iterator().cost());
    // multi-valued terms
    terms = new DummyTerms(42, 63);
    builder = new DocIdSetBuilder(100, terms);
    assertEquals(1.5, builder.numValuesPerDoc, 0d);
    assertTrue(builder.multivalued);
    adder = builder.grow(2);
    adder.add(5);
    adder.add(7);
    set = builder.build();
    assertTrue(set instanceof BitDocIdSet);
    // it thinks the same doc was added twice
    assertEquals(1, set.iterator().cost());
    // incomplete stats
    terms = new DummyTerms(42, -1);
    builder = new DocIdSetBuilder(100, terms);
    assertEquals(1d, builder.numValuesPerDoc, 0d);
    assertTrue(builder.multivalued);
    terms = new DummyTerms(-1, 84);
    builder = new DocIdSetBuilder(100, terms);
    assertEquals(1d, builder.numValuesPerDoc, 0d);
    assertTrue(builder.multivalued);
}

Example 25 with DocIdSet

use of org.apache.lucene.search.DocIdSet in project greplin-lucene-utils by Cue.

the class BaseFilterTest method assertFilterBitsEqual.

protected static void assertFilterBitsEqual(IndexReader reader, Filter filter, boolean... expected) throws IOException {
    DocIdSet actual = filter.getDocIdSet(reader);
    FixedBitSet actualBitSet = new FixedBitSet(expected.length);
    actualBitSet.or(actual.iterator());
    for (int i = 0; i < expected.length; i++) {
        Assert.assertEquals("Expected same value at position " + i, expected[i], actualBitSet.get(i));
    }
}