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

Example 66 with MatchAllDocsQuery

use of org.apache.lucene.search.MatchAllDocsQuery 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 67 with MatchAllDocsQuery

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

the class TestOrdinalMappingLeafReader method verifyResults.

private void verifyResults(Directory indexDir, Directory taxoDir) throws IOException {
    DirectoryReader indexReader = DirectoryReader.open(indexDir);
    DirectoryTaxonomyReader taxoReader = new DirectoryTaxonomyReader(taxoDir);
    IndexSearcher searcher = newSearcher(indexReader);
    FacetsCollector collector = new FacetsCollector();
    FacetsCollector.search(searcher, new MatchAllDocsQuery(), 10, collector);
    // tag facets
    Facets tagFacets = new FastTaxonomyFacetCounts("$tags", taxoReader, facetConfig, collector);
    FacetResult result = tagFacets.getTopChildren(10, "tag");
    for (LabelAndValue lv : result.labelValues) {
        if (VERBOSE) {
            System.out.println(lv);
        }
        assertEquals(NUM_DOCS, lv.value.intValue());
    }
    // id facets
    Facets idFacets = new FastTaxonomyFacetCounts(taxoReader, facetConfig, collector);
    FacetResult idResult = idFacets.getTopChildren(10, "id");
    assertEquals(NUM_DOCS, idResult.childCount);
    // each "id" appears twice
    assertEquals(NUM_DOCS * 2, idResult.value);
    BinaryDocValues bdv = MultiDocValues.getBinaryValues(indexReader, "bdv");
    BinaryDocValues cbdv = MultiDocValues.getBinaryValues(indexReader, "cbdv");
    for (int i = 0; i < indexReader.maxDoc(); i++) {
        assertEquals(i, bdv.nextDoc());
        assertEquals(i, cbdv.nextDoc());
        assertEquals(Integer.parseInt(cbdv.binaryValue().utf8ToString()), Integer.parseInt(bdv.binaryValue().utf8ToString()) * 2);
    }
    IOUtils.close(indexReader, taxoReader);
}

Example 68 with MatchAllDocsQuery

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

the class TestSearcherTaxonomyManager method testDirectory.

public void testDirectory() throws Exception {
    Directory indexDir = newDirectory();
    Directory taxoDir = newDirectory();
    final IndexWriter w = new IndexWriter(indexDir, newIndexWriterConfig(new MockAnalyzer(random())));
    final DirectoryTaxonomyWriter tw = new DirectoryTaxonomyWriter(taxoDir);
    // first empty commit
    w.commit();
    tw.commit();
    final SearcherTaxonomyManager mgr = new SearcherTaxonomyManager(indexDir, taxoDir, null);
    final FacetsConfig config = new FacetsConfig();
    config.setMultiValued("field", true);
    final AtomicBoolean stop = new AtomicBoolean();
    // How many unique facets to index before stopping:
    final int ordLimit = TEST_NIGHTLY ? 100000 : 6000;
    Thread indexer = new IndexerThread(w, config, tw, mgr, ordLimit, stop);
    indexer.start();
    try {
        while (!stop.get()) {
            SearcherAndTaxonomy pair = mgr.acquire();
            try {
                //System.out.println("search maxOrd=" + pair.taxonomyReader.getSize());
                FacetsCollector sfc = new FacetsCollector();
                pair.searcher.search(new MatchAllDocsQuery(), sfc);
                Facets facets = getTaxonomyFacetCounts(pair.taxonomyReader, config, sfc);
                FacetResult result = facets.getTopChildren(10, "field");
                if (pair.searcher.getIndexReader().numDocs() > 0) {
                    //System.out.println(pair.taxonomyReader.getSize());
                    assertTrue(result.childCount > 0);
                    assertTrue(result.labelValues.length > 0);
                }
            //if (VERBOSE) {
            //System.out.println("TEST: facets=" + FacetTestUtils.toString(results.get(0)));
            //}
            } finally {
                mgr.release(pair);
            }
        }
    } finally {
        indexer.join();
    }
    if (VERBOSE) {
        System.out.println("TEST: now stop");
    }
    w.close();
    IOUtils.close(mgr, tw, taxoDir, indexDir);
}

Example 69 with MatchAllDocsQuery

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

the class TestRangeFacetCounts method testCustomDoubleValuesSource.

public void testCustomDoubleValuesSource() throws Exception {
    Directory dir = newDirectory();
    RandomIndexWriter writer = new RandomIndexWriter(random(), dir);
    Document doc = new Document();
    writer.addDocument(doc);
    writer.addDocument(doc);
    writer.addDocument(doc);
    // Test wants 3 docs in one segment:
    writer.forceMerge(1);
    final DoubleValuesSource vs = new DoubleValuesSource() {

        @Override
        public DoubleValues getValues(LeafReaderContext ctx, DoubleValues scores) throws IOException {
            return new DoubleValues() {

                int doc = -1;

                @Override
                public double doubleValue() throws IOException {
                    return doc + 1;
                }

                @Override
                public boolean advanceExact(int doc) throws IOException {
                    this.doc = doc;
                    return true;
                }
            };
        }

        @Override
        public boolean needsScores() {
            return false;
        }

        @Override
        public Explanation explain(LeafReaderContext ctx, int docId, Explanation scoreExplanation) throws IOException {
            return Explanation.match(docId + 1, "");
        }
    };
    FacetsConfig config = new FacetsConfig();
    FacetsCollector fc = new FacetsCollector();
    IndexReader r = writer.getReader();
    IndexSearcher s = newSearcher(r);
    s.search(new MatchAllDocsQuery(), fc);
    final DoubleRange[] ranges = new DoubleRange[] { new DoubleRange("< 1", 0.0, true, 1.0, false), new DoubleRange("< 2", 0.0, true, 2.0, false), new DoubleRange("< 5", 0.0, true, 5.0, false), new DoubleRange("< 10", 0.0, true, 10.0, false), new DoubleRange("< 20", 0.0, true, 20.0, false), new DoubleRange("< 50", 0.0, true, 50.0, false) };
    final Query fastMatchFilter;
    final AtomicBoolean filterWasUsed = new AtomicBoolean();
    if (random().nextBoolean()) {
        // Sort of silly:
        final Query in = new MatchAllDocsQuery();
        fastMatchFilter = new UsedQuery(in, filterWasUsed);
    } else {
        fastMatchFilter = null;
    }
    if (VERBOSE) {
        System.out.println("TEST: fastMatchFilter=" + fastMatchFilter);
    }
    Facets facets = new DoubleRangeFacetCounts("field", vs, fc, fastMatchFilter, ranges);
    assertEquals("dim=field path=[] value=3 childCount=6\n  < 1 (0)\n  < 2 (1)\n  < 5 (3)\n  < 10 (3)\n  < 20 (3)\n  < 50 (3)\n", facets.getTopChildren(10, "field").toString());
    assertTrue(fastMatchFilter == null || filterWasUsed.get());
    DrillDownQuery ddq = new DrillDownQuery(config);
    ddq.add("field", ranges[1].getQuery(fastMatchFilter, vs));
    // Test simple drill-down:
    assertEquals(1, s.search(ddq, 10).totalHits);
    // Test drill-sideways after drill-down
    DrillSideways ds = new DrillSideways(s, config, (TaxonomyReader) null) {

        @Override
        protected Facets buildFacetsResult(FacetsCollector drillDowns, FacetsCollector[] drillSideways, String[] drillSidewaysDims) throws IOException {
            assert drillSideways.length == 1;
            return new DoubleRangeFacetCounts("field", vs, drillSideways[0], fastMatchFilter, ranges);
        }

        @Override
        protected boolean scoreSubDocsAtOnce() {
            return random().nextBoolean();
        }
    };
    DrillSidewaysResult dsr = ds.search(ddq, 10);
    assertEquals(1, dsr.hits.totalHits);
    assertEquals("dim=field path=[] value=3 childCount=6\n  < 1 (0)\n  < 2 (1)\n  < 5 (3)\n  < 10 (3)\n  < 20 (3)\n  < 50 (3)\n", dsr.facets.getTopChildren(10, "field").toString());
    writer.close();
    IOUtils.close(r, dir);
}

Example 70 with MatchAllDocsQuery

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

the class TestTaxonomyFacetCounts method testWrongIndexFieldName.

public void testWrongIndexFieldName() throws Exception {
    Directory dir = newDirectory();
    Directory taxoDir = newDirectory();
    // Writes facet ords to a separate directory from the
    // main index:
    DirectoryTaxonomyWriter taxoWriter = new DirectoryTaxonomyWriter(taxoDir, IndexWriterConfig.OpenMode.CREATE);
    FacetsConfig config = new FacetsConfig();
    config.setIndexFieldName("a", "$facets2");
    RandomIndexWriter writer = new RandomIndexWriter(random(), dir);
    Document doc = new Document();
    doc.add(new FacetField("a", "foo1"));
    writer.addDocument(config.build(taxoWriter, doc));
    // NRT open
    IndexSearcher searcher = newSearcher(writer.getReader());
    // NRT open
    TaxonomyReader taxoReader = new DirectoryTaxonomyReader(taxoWriter);
    FacetsCollector c = new FacetsCollector();
    searcher.search(new MatchAllDocsQuery(), c);
    // Uses default $facets field:
    Facets facets;
    if (random().nextBoolean()) {
        facets = new FastTaxonomyFacetCounts(taxoReader, config, c);
    } else {
        OrdinalsReader ordsReader = new DocValuesOrdinalsReader();
        if (random().nextBoolean()) {
            ordsReader = new CachedOrdinalsReader(ordsReader);
        }
        facets = new TaxonomyFacetCounts(ordsReader, taxoReader, config, c);
    }
    // Ask for top 10 labels for any dims that have counts:
    List<FacetResult> results = facets.getAllDims(10);
    assertTrue(results.isEmpty());
    expectThrows(IllegalArgumentException.class, () -> {
        facets.getSpecificValue("a");
    });
    expectThrows(IllegalArgumentException.class, () -> {
        facets.getTopChildren(10, "a");
    });
    writer.close();
    IOUtils.close(taxoWriter, searcher.getIndexReader(), taxoReader, taxoDir, dir);
}