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

Example 26 with DocIdSetIterator

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

the class TestFixedBitSet method doIterate2.

void doIterate2(java.util.BitSet a, FixedBitSet b) throws IOException {
    assertEquals(a.cardinality(), b.cardinality());
    int aa = -1, bb = -1;
    DocIdSetIterator iterator = new BitSetIterator(b, 0);
    do {
        aa = a.nextSetBit(aa + 1);
        bb = random().nextBoolean() ? iterator.nextDoc() : iterator.advance(bb + 1);
        assertEquals(aa == -1 ? DocIdSetIterator.NO_MORE_DOCS : aa, bb);
    } while (aa >= 0);
}

Example 27 with DocIdSetIterator

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

the class TestDrillSideways method testRandom.

public void testRandom() throws Exception {
    while (aChance == 0.0) {
        aChance = random().nextDouble();
    }
    while (bChance == 0.0) {
        bChance = random().nextDouble();
    }
    while (cChance == 0.0) {
        cChance = random().nextDouble();
    }
    //aChance = .01;
    //bChance = 0.5;
    //cChance = 1.0;
    double sum = aChance + bChance + cChance;
    aChance /= sum;
    bChance /= sum;
    cChance /= sum;
    int numDims = TestUtil.nextInt(random(), 2, 5);
    //int numDims = 3;
    int numDocs = atLeast(3000);
    //int numDocs = 20;
    if (VERBOSE) {
        System.out.println("numDims=" + numDims + " numDocs=" + numDocs + " aChance=" + aChance + " bChance=" + bChance + " cChance=" + cChance);
    }
    String[][] dimValues = new String[numDims][];
    int valueCount = 2;
    for (int dim = 0; dim < numDims; dim++) {
        Set<String> values = new HashSet<>();
        while (values.size() < valueCount) {
            String s = TestUtil.randomRealisticUnicodeString(random());
            //String s = _TestUtil.randomString(random());
            if (s.length() > 0) {
                values.add(s);
            }
        }
        dimValues[dim] = values.toArray(new String[values.size()]);
        valueCount *= 2;
    }
    List<Doc> docs = new ArrayList<>();
    for (int i = 0; i < numDocs; i++) {
        Doc doc = new Doc();
        doc.id = "" + i;
        doc.contentToken = randomContentToken(false);
        doc.dims = new int[numDims];
        doc.dims2 = new int[numDims];
        for (int dim = 0; dim < numDims; dim++) {
            if (random().nextInt(5) == 3) {
                // This doc is missing this dim:
                doc.dims[dim] = -1;
            } else if (dimValues[dim].length <= 4) {
                int dimUpto = 0;
                doc.dims[dim] = dimValues[dim].length - 1;
                while (dimUpto < dimValues[dim].length) {
                    if (random().nextBoolean()) {
                        doc.dims[dim] = dimUpto;
                        break;
                    }
                    dimUpto++;
                }
            } else {
                doc.dims[dim] = random().nextInt(dimValues[dim].length);
            }
            if (random().nextInt(5) == 3) {
                // 2nd value:
                doc.dims2[dim] = random().nextInt(dimValues[dim].length);
            } else {
                doc.dims2[dim] = -1;
            }
        }
        docs.add(doc);
    }
    Directory d = newDirectory();
    Directory td = newDirectory();
    IndexWriterConfig iwc = newIndexWriterConfig(new MockAnalyzer(random()));
    iwc.setInfoStream(InfoStream.NO_OUTPUT);
    RandomIndexWriter w = new RandomIndexWriter(random(), d, iwc);
    DirectoryTaxonomyWriter tw = new DirectoryTaxonomyWriter(td, IndexWriterConfig.OpenMode.CREATE);
    FacetsConfig config = new FacetsConfig();
    for (int i = 0; i < numDims; i++) {
        config.setMultiValued("dim" + i, true);
    }
    boolean doUseDV = random().nextBoolean();
    for (Doc rawDoc : docs) {
        Document doc = new Document();
        doc.add(newStringField("id", rawDoc.id, Field.Store.YES));
        doc.add(new SortedDocValuesField("id", new BytesRef(rawDoc.id)));
        doc.add(newStringField("content", rawDoc.contentToken, Field.Store.NO));
        if (VERBOSE) {
            System.out.println("  doc id=" + rawDoc.id + " token=" + rawDoc.contentToken);
        }
        for (int dim = 0; dim < numDims; dim++) {
            int dimValue = rawDoc.dims[dim];
            if (dimValue != -1) {
                if (doUseDV) {
                    doc.add(new SortedSetDocValuesFacetField("dim" + dim, dimValues[dim][dimValue]));
                } else {
                    doc.add(new FacetField("dim" + dim, dimValues[dim][dimValue]));
                }
                doc.add(new StringField("dim" + dim, dimValues[dim][dimValue], Field.Store.YES));
                if (VERBOSE) {
                    System.out.println("    dim" + dim + "=" + new BytesRef(dimValues[dim][dimValue]));
                }
            }
            int dimValue2 = rawDoc.dims2[dim];
            if (dimValue2 != -1) {
                if (doUseDV) {
                    doc.add(new SortedSetDocValuesFacetField("dim" + dim, dimValues[dim][dimValue2]));
                } else {
                    doc.add(new FacetField("dim" + dim, dimValues[dim][dimValue2]));
                }
                doc.add(new StringField("dim" + dim, dimValues[dim][dimValue2], Field.Store.YES));
                if (VERBOSE) {
                    System.out.println("      dim" + dim + "=" + new BytesRef(dimValues[dim][dimValue2]));
                }
            }
        }
        w.addDocument(config.build(tw, doc));
    }
    if (random().nextBoolean()) {
        // Randomly delete a few docs:
        int numDel = TestUtil.nextInt(random(), 1, (int) (numDocs * 0.05));
        if (VERBOSE) {
            System.out.println("delete " + numDel);
        }
        int delCount = 0;
        while (delCount < numDel) {
            Doc doc = docs.get(random().nextInt(docs.size()));
            if (!doc.deleted) {
                if (VERBOSE) {
                    System.out.println("  delete id=" + doc.id);
                }
                doc.deleted = true;
                w.deleteDocuments(new Term("id", doc.id));
                delCount++;
            }
        }
    }
    if (random().nextBoolean()) {
        if (VERBOSE) {
            System.out.println("TEST: forceMerge(1)...");
        }
        w.forceMerge(1);
    }
    IndexReader r = w.getReader();
    final SortedSetDocValuesReaderState sortedSetDVState;
    IndexSearcher s = newSearcher(r);
    if (doUseDV) {
        sortedSetDVState = new DefaultSortedSetDocValuesReaderState(s.getIndexReader());
    } else {
        sortedSetDVState = null;
    }
    if (VERBOSE) {
        System.out.println("r.numDocs() = " + r.numDocs());
    }
    // NRT open
    TaxonomyReader tr = new DirectoryTaxonomyReader(tw);
    int numIters = atLeast(10);
    for (int iter = 0; iter < numIters; iter++) {
        String contentToken = random().nextInt(30) == 17 ? null : randomContentToken(true);
        int numDrillDown = TestUtil.nextInt(random(), 1, Math.min(4, numDims));
        if (VERBOSE) {
            System.out.println("\nTEST: iter=" + iter + " baseQuery=" + contentToken + " numDrillDown=" + numDrillDown + " useSortedSetDV=" + doUseDV);
        }
        String[][] drillDowns = new String[numDims][];
        int count = 0;
        boolean anyMultiValuedDrillDowns = false;
        while (count < numDrillDown) {
            int dim = random().nextInt(numDims);
            if (drillDowns[dim] == null) {
                if (random().nextBoolean()) {
                    // Drill down on one value:
                    drillDowns[dim] = new String[] { dimValues[dim][random().nextInt(dimValues[dim].length)] };
                } else {
                    int orCount = TestUtil.nextInt(random(), 1, Math.min(5, dimValues[dim].length));
                    drillDowns[dim] = new String[orCount];
                    anyMultiValuedDrillDowns |= orCount > 1;
                    for (int i = 0; i < orCount; i++) {
                        while (true) {
                            String value = dimValues[dim][random().nextInt(dimValues[dim].length)];
                            for (int j = 0; j < i; j++) {
                                if (value.equals(drillDowns[dim][j])) {
                                    value = null;
                                    break;
                                }
                            }
                            if (value != null) {
                                drillDowns[dim][i] = value;
                                break;
                            }
                        }
                    }
                }
                if (VERBOSE) {
                    BytesRef[] values = new BytesRef[drillDowns[dim].length];
                    for (int i = 0; i < values.length; i++) {
                        values[i] = new BytesRef(drillDowns[dim][i]);
                    }
                    System.out.println("  dim" + dim + "=" + Arrays.toString(values));
                }
                count++;
            }
        }
        Query baseQuery;
        if (contentToken == null) {
            baseQuery = new MatchAllDocsQuery();
        } else {
            baseQuery = new TermQuery(new Term("content", contentToken));
        }
        DrillDownQuery ddq = new DrillDownQuery(config, baseQuery);
        for (int dim = 0; dim < numDims; dim++) {
            if (drillDowns[dim] != null) {
                for (String value : drillDowns[dim]) {
                    ddq.add("dim" + dim, value);
                }
            }
        }
        Query filter;
        if (random().nextInt(7) == 6) {
            if (VERBOSE) {
                System.out.println("  only-even filter");
            }
            filter = new Query() {

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

                        @Override
                        public Scorer scorer(LeafReaderContext context) throws IOException {
                            DocIdSetIterator approximation = DocIdSetIterator.all(context.reader().maxDoc());
                            return new ConstantScoreScorer(this, score(), new TwoPhaseIterator(approximation) {

                                @Override
                                public boolean matches() throws IOException {
                                    int docID = approximation.docID();
                                    return (Integer.parseInt(context.reader().document(docID).get("id")) & 1) == 0;
                                }

                                @Override
                                public float matchCost() {
                                    return 1000f;
                                }
                            });
                        }
                    };
                }

                @Override
                public String toString(String field) {
                    return "drillSidewaysTestFilter";
                }

                @Override
                public boolean equals(Object o) {
                    return o == this;
                }

                @Override
                public int hashCode() {
                    return System.identityHashCode(this);
                }
            };
        } else {
            filter = null;
        }
        // Verify docs are always collected in order.  If we
        // had an AssertingScorer it could catch it when
        // Weight.scoresDocsOutOfOrder lies!:
        getNewDrillSideways(s, config, tr).search(ddq, new SimpleCollector() {

            int lastDocID;

            @Override
            public void collect(int doc) {
                assert doc > lastDocID;
                lastDocID = doc;
            }

            @Override
            protected void doSetNextReader(LeafReaderContext context) throws IOException {
                lastDocID = -1;
            }

            @Override
            public boolean needsScores() {
                return false;
            }
        });
        // subScorers are on the same docID:
        if (!anyMultiValuedDrillDowns) {
            // Can only do this test when there are no OR'd
            // drill-down values, because in that case it's
            // easily possible for one of the DD terms to be on
            // a future docID:
            getNewDrillSidewaysScoreSubdocsAtOnce(s, config, tr).search(ddq, new AssertingSubDocsAtOnceCollector());
        }
        TestFacetResult expected = slowDrillSidewaysSearch(s, docs, contentToken, drillDowns, dimValues, filter);
        Sort sort = new Sort(new SortField("id", SortField.Type.STRING));
        DrillSideways ds;
        if (doUseDV) {
            ds = getNewDrillSideways(s, config, sortedSetDVState);
        } else {
            ds = getNewDrillSidewaysBuildFacetsResult(s, config, tr);
        }
        // Retrieve all facets:
        DrillSidewaysResult actual = ds.search(ddq, filter, null, numDocs, sort, true, true);
        TopDocs hits = s.search(baseQuery, numDocs);
        Map<String, Float> scores = new HashMap<>();
        for (ScoreDoc sd : hits.scoreDocs) {
            scores.put(s.doc(sd.doc).get("id"), sd.score);
        }
        if (VERBOSE) {
            System.out.println("  verify all facets");
        }
        verifyEquals(dimValues, s, expected, actual, scores, doUseDV);
        // Make sure drill down doesn't change score:
        Query q = ddq;
        if (filter != null) {
            q = new BooleanQuery.Builder().add(q, Occur.MUST).add(filter, Occur.FILTER).build();
        }
        TopDocs ddqHits = s.search(q, numDocs);
        assertEquals(expected.hits.size(), ddqHits.totalHits);
        for (int i = 0; i < expected.hits.size(); i++) {
            // Score should be IDENTICAL:
            assertEquals(scores.get(expected.hits.get(i).id), ddqHits.scoreDocs[i].score, 0.0f);
        }
    }
    w.close();
    IOUtils.close(r, tr, tw, d, td);
}

Example 28 with DocIdSetIterator

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

the class TestUnifiedHighlighterExtensibility method testUnifiedHighlighterExtensibility.

/**
   * This test is for maintaining the extensibility of the UnifiedHighlighter
   * for customizations out of package.
   */
@Test
public void testUnifiedHighlighterExtensibility() {
    final int maxLength = 1000;
    UnifiedHighlighter uh = new UnifiedHighlighter(null, new MockAnalyzer(random())) {

        @Override
        protected Map<String, Object[]> highlightFieldsAsObjects(String[] fieldsIn, Query query, int[] docIdsIn, int[] maxPassagesIn) throws IOException {
            return super.highlightFieldsAsObjects(fieldsIn, query, docIdsIn, maxPassagesIn);
        }

        @Override
        protected OffsetSource getOffsetSource(String field) {
            return super.getOffsetSource(field);
        }

        @Override
        protected BreakIterator getBreakIterator(String field) {
            return super.getBreakIterator(field);
        }

        @Override
        protected PassageScorer getScorer(String field) {
            return super.getScorer(field);
        }

        @Override
        protected PassageFormatter getFormatter(String field) {
            return super.getFormatter(field);
        }

        @Override
        public Analyzer getIndexAnalyzer() {
            return super.getIndexAnalyzer();
        }

        @Override
        public IndexSearcher getIndexSearcher() {
            return super.getIndexSearcher();
        }

        @Override
        protected int getMaxNoHighlightPassages(String field) {
            return super.getMaxNoHighlightPassages(field);
        }

        @Override
        protected Boolean requiresRewrite(SpanQuery spanQuery) {
            return super.requiresRewrite(spanQuery);
        }

        @Override
        protected LimitedStoredFieldVisitor newLimitedStoredFieldsVisitor(String[] fields) {
            return super.newLimitedStoredFieldsVisitor(fields);
        }

        @Override
        protected List<CharSequence[]> loadFieldValues(String[] fields, DocIdSetIterator docIter, int cacheCharsThreshold) throws IOException {
            return super.loadFieldValues(fields, docIter, cacheCharsThreshold);
        }

        @Override
        protected FieldHighlighter getFieldHighlighter(String field, Query query, Set<Term> allTerms, int maxPassages) {
            // THIS IS A COPY of the superclass impl; but use CustomFieldHighlighter
            BytesRef[] terms = filterExtractedTerms(getFieldMatcher(field), allTerms);
            Set<HighlightFlag> highlightFlags = getFlags(field);
            PhraseHelper phraseHelper = getPhraseHelper(field, query, highlightFlags);
            CharacterRunAutomaton[] automata = getAutomata(field, query, highlightFlags);
            OffsetSource offsetSource = getOptimizedOffsetSource(field, terms, phraseHelper, automata);
            return new CustomFieldHighlighter(field, getOffsetStrategy(offsetSource, field, terms, phraseHelper, automata, highlightFlags), new SplittingBreakIterator(getBreakIterator(field), UnifiedHighlighter.MULTIVAL_SEP_CHAR), getScorer(field), maxPassages, getMaxNoHighlightPassages(field), getFormatter(field));
        }

        @Override
        protected FieldOffsetStrategy getOffsetStrategy(OffsetSource offsetSource, String field, BytesRef[] terms, PhraseHelper phraseHelper, CharacterRunAutomaton[] automata, Set<HighlightFlag> highlightFlags) {
            return super.getOffsetStrategy(offsetSource, field, terms, phraseHelper, automata, highlightFlags);
        }

        @Override
        public int getMaxLength() {
            return maxLength;
        }
    };
    assertEquals(uh.getMaxLength(), maxLength);
}

Example 29 with DocIdSetIterator

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

the class BaseBitSetTestCase method testOr.

private void testOr(float load) throws IOException {
    final int numBits = 1 + random().nextInt(100000);
    // empty
    BitSet set1 = new JavaUtilBitSet(randomSet(numBits, 0), numBits);
    T set2 = copyOf(set1, numBits);
    final int iterations = atLeast(10);
    for (int iter = 0; iter < iterations; ++iter) {
        DocIdSet otherSet = randomCopy(new JavaUtilBitSet(randomSet(numBits, load), numBits), numBits);
        DocIdSetIterator otherIterator = otherSet.iterator();
        if (otherIterator != null) {
            set1.or(otherIterator);
            set2.or(otherSet.iterator());
            assertEquals(set1, set2, numBits);
        }
    }
}

Example 30 with DocIdSetIterator

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

the class ExportWriter method writeDocs.

protected void writeDocs(SolrQueryRequest req, IteratorWriter.ItemWriter writer, Sort sort) throws IOException {
    //Write the data.
    List<LeafReaderContext> leaves = req.getSearcher().getTopReaderContext().leaves();
    SortDoc sortDoc = getSortDoc(req.getSearcher(), sort.getSort());
    int count = 0;
    int queueSize = 30000;
    SortQueue queue = new SortQueue(queueSize, sortDoc);
    SortDoc[] outDocs = new SortDoc[queueSize];
    while (count < totalHits) {
        //long begin = System.nanoTime();
        queue.reset();
        SortDoc top = queue.top();
        for (int i = 0; i < leaves.size(); i++) {
            sortDoc.setNextReader(leaves.get(i));
            // cost is not useful here
            DocIdSetIterator it = new BitSetIterator(sets[i], 0);
            int docId = -1;
            while ((docId = it.nextDoc()) != DocIdSetIterator.NO_MORE_DOCS) {
                sortDoc.setValues(docId);
                if (top.lessThan(sortDoc)) {
                    top.setValues(sortDoc);
                    top = queue.updateTop();
                }
            }
        }
        int outDocsIndex = -1;
        for (int i = 0; i < queueSize; i++) {
            SortDoc s = queue.pop();
            if (s.docId > -1) {
                outDocs[++outDocsIndex] = s;
            }
        }
        //long end = System.nanoTime();
        count += (outDocsIndex + 1);
        try {
            for (int i = outDocsIndex; i >= 0; --i) {
                SortDoc s = outDocs[i];
                writer.add((MapWriter) ew -> {
                    writeDoc(s, leaves, ew);
                    s.reset();
                });
            }
        } catch (Throwable e) {
            Throwable ex = e;
            while (ex != null) {
                String m = ex.getMessage();
                if (m != null && m.contains("Broken pipe")) {
                    throw new IgnoreException();
                }
                ex = ex.getCause();
            }
            if (e instanceof IOException) {
                throw ((IOException) e);
            } else {
                throw new IOException(e);
            }
        }
    }
}