Examples with FacetsConfig org.apache.lucene.facet.FacetsConfig used on opensource projects

Example 31 with FacetsConfig

use of org.apache.lucene.facet.FacetsConfig in project lucene-solr by apache.

the class TestDirectoryTaxonomyWriter method testHugeLabel.

@Test
public void testHugeLabel() throws Exception {
    Directory indexDir = newDirectory(), taxoDir = newDirectory();
    IndexWriter indexWriter = new IndexWriter(indexDir, newIndexWriterConfig(new MockAnalyzer(random())));
    DirectoryTaxonomyWriter taxoWriter = new DirectoryTaxonomyWriter(taxoDir, OpenMode.CREATE, new Cl2oTaxonomyWriterCache(2, 1f, 1));
    FacetsConfig config = new FacetsConfig();
    // Add one huge label:
    String bigs = null;
    int ordinal = -1;
    // for the dimension and separator
    int len = FacetLabel.MAX_CATEGORY_PATH_LENGTH - 4;
    bigs = TestUtil.randomSimpleString(random(), len, len);
    FacetField ff = new FacetField("dim", bigs);
    FacetLabel cp = new FacetLabel("dim", bigs);
    ordinal = taxoWriter.addCategory(cp);
    Document doc = new Document();
    doc.add(ff);
    indexWriter.addDocument(config.build(taxoWriter, doc));
    // Add tiny ones to cause a re-hash
    for (int i = 0; i < 3; i++) {
        String s = TestUtil.randomSimpleString(random(), 1, 10);
        taxoWriter.addCategory(new FacetLabel("dim", s));
        doc = new Document();
        doc.add(new FacetField("dim", s));
        indexWriter.addDocument(config.build(taxoWriter, doc));
    }
    // when too large components were allowed to be added, this resulted in a new added category
    assertEquals(ordinal, taxoWriter.addCategory(cp));
    indexWriter.close();
    IOUtils.close(taxoWriter);
    DirectoryReader indexReader = DirectoryReader.open(indexDir);
    TaxonomyReader taxoReader = new DirectoryTaxonomyReader(taxoDir);
    IndexSearcher searcher = new IndexSearcher(indexReader);
    DrillDownQuery ddq = new DrillDownQuery(new FacetsConfig());
    ddq.add("dim", bigs);
    assertEquals(1, searcher.search(ddq, 10).totalHits);
    IOUtils.close(indexReader, taxoReader, indexDir, taxoDir);
}

Example 32 with FacetsConfig

use of org.apache.lucene.facet.FacetsConfig in project lucene-solr by apache.

the class TestRangeFacetCounts method testRandomLongs.

public void testRandomLongs() throws Exception {
    Directory dir = newDirectory();
    RandomIndexWriter w = new RandomIndexWriter(random(), dir);
    int numDocs = atLeast(1000);
    if (VERBOSE) {
        System.out.println("TEST: numDocs=" + numDocs);
    }
    long[] values = new long[numDocs];
    long minValue = Long.MAX_VALUE;
    long maxValue = Long.MIN_VALUE;
    for (int i = 0; i < numDocs; i++) {
        Document doc = new Document();
        long v = random().nextLong();
        values[i] = v;
        doc.add(new NumericDocValuesField("field", v));
        doc.add(new LongPoint("field", v));
        w.addDocument(doc);
        minValue = Math.min(minValue, v);
        maxValue = Math.max(maxValue, v);
    }
    IndexReader r = w.getReader();
    IndexSearcher s = newSearcher(r, false);
    FacetsConfig config = new FacetsConfig();
    int numIters = atLeast(10);
    for (int iter = 0; iter < numIters; iter++) {
        if (VERBOSE) {
            System.out.println("TEST: iter=" + iter);
        }
        int numRange = TestUtil.nextInt(random(), 1, 100);
        LongRange[] ranges = new LongRange[numRange];
        int[] expectedCounts = new int[numRange];
        long minAcceptedValue = Long.MAX_VALUE;
        long maxAcceptedValue = Long.MIN_VALUE;
        for (int rangeID = 0; rangeID < numRange; rangeID++) {
            long min;
            if (rangeID > 0 && random().nextInt(10) == 7) {
                // Use an existing boundary:
                LongRange prevRange = ranges[random().nextInt(rangeID)];
                if (random().nextBoolean()) {
                    min = prevRange.min;
                } else {
                    min = prevRange.max;
                }
            } else {
                min = random().nextLong();
            }
            long max;
            if (rangeID > 0 && random().nextInt(10) == 7) {
                // Use an existing boundary:
                LongRange prevRange = ranges[random().nextInt(rangeID)];
                if (random().nextBoolean()) {
                    max = prevRange.min;
                } else {
                    max = prevRange.max;
                }
            } else {
                max = random().nextLong();
            }
            if (min > max) {
                long x = min;
                min = max;
                max = x;
            }
            boolean minIncl;
            boolean maxIncl;
            // NOTE: max - min >= 0 is here to handle the common overflow case!
            if (max - min >= 0 && max - min < 2) {
                // If max == min or max == min+1, we always do inclusive, else we might pass an empty range and hit exc from LongRange's ctor:
                minIncl = true;
                maxIncl = true;
            } else {
                minIncl = random().nextBoolean();
                maxIncl = random().nextBoolean();
            }
            ranges[rangeID] = new LongRange("r" + rangeID, min, minIncl, max, maxIncl);
            if (VERBOSE) {
                System.out.println("  range " + rangeID + ": " + ranges[rangeID]);
            }
            // expected count:
            for (int i = 0; i < numDocs; i++) {
                boolean accept = true;
                if (minIncl) {
                    accept &= values[i] >= min;
                } else {
                    accept &= values[i] > min;
                }
                if (maxIncl) {
                    accept &= values[i] <= max;
                } else {
                    accept &= values[i] < max;
                }
                if (accept) {
                    expectedCounts[rangeID]++;
                    minAcceptedValue = Math.min(minAcceptedValue, values[i]);
                    maxAcceptedValue = Math.max(maxAcceptedValue, values[i]);
                }
            }
        }
        FacetsCollector sfc = new FacetsCollector();
        s.search(new MatchAllDocsQuery(), sfc);
        Query fastMatchQuery;
        if (random().nextBoolean()) {
            if (random().nextBoolean()) {
                fastMatchQuery = LongPoint.newRangeQuery("field", minValue, maxValue);
            } else {
                fastMatchQuery = LongPoint.newRangeQuery("field", minAcceptedValue, maxAcceptedValue);
            }
        } else {
            fastMatchQuery = null;
        }
        LongValuesSource vs = LongValuesSource.fromLongField("field");
        Facets facets = new LongRangeFacetCounts("field", vs, sfc, fastMatchQuery, ranges);
        FacetResult result = facets.getTopChildren(10, "field");
        assertEquals(numRange, result.labelValues.length);
        for (int rangeID = 0; rangeID < numRange; rangeID++) {
            if (VERBOSE) {
                System.out.println("  range " + rangeID + " expectedCount=" + expectedCounts[rangeID]);
            }
            LabelAndValue subNode = result.labelValues[rangeID];
            assertEquals("r" + rangeID, subNode.label);
            assertEquals(expectedCounts[rangeID], subNode.value.intValue());
            LongRange range = ranges[rangeID];
            // Test drill-down:
            DrillDownQuery ddq = new DrillDownQuery(config);
            if (random().nextBoolean()) {
                ddq.add("field", LongPoint.newRangeQuery("field", range.min, range.max));
            } else {
                ddq.add("field", range.getQuery(fastMatchQuery, vs));
            }
            assertEquals(expectedCounts[rangeID], s.search(ddq, 10).totalHits);
        }
    }
    w.close();
    IOUtils.close(r, dir);
}

Example 33 with FacetsConfig

use of org.apache.lucene.facet.FacetsConfig in project lucene-solr by apache.

the class TestRangeFacetCounts method testRandomDoubles.

public void testRandomDoubles() throws Exception {
    Directory dir = newDirectory();
    RandomIndexWriter w = new RandomIndexWriter(random(), dir);
    int numDocs = atLeast(1000);
    double[] values = new double[numDocs];
    double minValue = Double.POSITIVE_INFINITY;
    double maxValue = Double.NEGATIVE_INFINITY;
    for (int i = 0; i < numDocs; i++) {
        Document doc = new Document();
        double v = random().nextDouble();
        values[i] = v;
        doc.add(new DoubleDocValuesField("field", v));
        doc.add(new DoublePoint("field", v));
        w.addDocument(doc);
        minValue = Math.min(minValue, v);
        maxValue = Math.max(maxValue, v);
    }
    IndexReader r = w.getReader();
    IndexSearcher s = newSearcher(r, false);
    FacetsConfig config = new FacetsConfig();
    int numIters = atLeast(10);
    for (int iter = 0; iter < numIters; iter++) {
        if (VERBOSE) {
            System.out.println("TEST: iter=" + iter);
        }
        int numRange = TestUtil.nextInt(random(), 1, 5);
        DoubleRange[] ranges = new DoubleRange[numRange];
        int[] expectedCounts = new int[numRange];
        double minAcceptedValue = Double.POSITIVE_INFINITY;
        double maxAcceptedValue = Double.NEGATIVE_INFINITY;
        for (int rangeID = 0; rangeID < numRange; rangeID++) {
            double min;
            if (rangeID > 0 && random().nextInt(10) == 7) {
                // Use an existing boundary:
                DoubleRange prevRange = ranges[random().nextInt(rangeID)];
                if (random().nextBoolean()) {
                    min = prevRange.min;
                } else {
                    min = prevRange.max;
                }
            } else {
                min = random().nextDouble();
            }
            double max;
            if (rangeID > 0 && random().nextInt(10) == 7) {
                // Use an existing boundary:
                DoubleRange prevRange = ranges[random().nextInt(rangeID)];
                if (random().nextBoolean()) {
                    max = prevRange.min;
                } else {
                    max = prevRange.max;
                }
            } else {
                max = random().nextDouble();
            }
            if (min > max) {
                double x = min;
                min = max;
                max = x;
            }
            boolean minIncl;
            boolean maxIncl;
            long minAsLong = NumericUtils.doubleToSortableLong(min);
            long maxAsLong = NumericUtils.doubleToSortableLong(max);
            // NOTE: maxAsLong - minAsLong >= 0 is here to handle the common overflow case!
            if (maxAsLong - minAsLong >= 0 && maxAsLong - minAsLong < 2) {
                minIncl = true;
                maxIncl = true;
            } else {
                minIncl = random().nextBoolean();
                maxIncl = random().nextBoolean();
            }
            ranges[rangeID] = new DoubleRange("r" + rangeID, min, minIncl, max, maxIncl);
            // expected count:
            for (int i = 0; i < numDocs; i++) {
                boolean accept = true;
                if (minIncl) {
                    accept &= values[i] >= min;
                } else {
                    accept &= values[i] > min;
                }
                if (maxIncl) {
                    accept &= values[i] <= max;
                } else {
                    accept &= values[i] < max;
                }
                if (accept) {
                    expectedCounts[rangeID]++;
                    minAcceptedValue = Math.min(minAcceptedValue, values[i]);
                    maxAcceptedValue = Math.max(maxAcceptedValue, values[i]);
                }
            }
        }
        FacetsCollector sfc = new FacetsCollector();
        s.search(new MatchAllDocsQuery(), sfc);
        Query fastMatchFilter;
        if (random().nextBoolean()) {
            if (random().nextBoolean()) {
                fastMatchFilter = DoublePoint.newRangeQuery("field", minValue, maxValue);
            } else {
                fastMatchFilter = DoublePoint.newRangeQuery("field", minAcceptedValue, maxAcceptedValue);
            }
        } else {
            fastMatchFilter = null;
        }
        DoubleValuesSource vs = DoubleValuesSource.fromDoubleField("field");
        Facets facets = new DoubleRangeFacetCounts("field", vs, sfc, fastMatchFilter, ranges);
        FacetResult result = facets.getTopChildren(10, "field");
        assertEquals(numRange, result.labelValues.length);
        for (int rangeID = 0; rangeID < numRange; rangeID++) {
            if (VERBOSE) {
                System.out.println("  range " + rangeID + " expectedCount=" + expectedCounts[rangeID]);
            }
            LabelAndValue subNode = result.labelValues[rangeID];
            assertEquals("r" + rangeID, subNode.label);
            assertEquals(expectedCounts[rangeID], subNode.value.intValue());
            DoubleRange range = ranges[rangeID];
            // Test drill-down:
            DrillDownQuery ddq = new DrillDownQuery(config);
            if (random().nextBoolean()) {
                ddq.add("field", DoublePoint.newRangeQuery("field", range.min, range.max));
            } else {
                ddq.add("field", range.getQuery(fastMatchFilter, vs));
            }
            assertEquals(expectedCounts[rangeID], s.search(ddq, 10).totalHits);
        }
    }
    w.close();
    IOUtils.close(r, dir);
}

Example 34 with FacetsConfig

use of org.apache.lucene.facet.FacetsConfig in project lucene-solr by apache.

the class TestRangeFacetCounts method testMixedRangeAndNonRangeTaxonomy.

/** Tests single request that mixes Range and non-Range
   *  faceting, with DrillSideways and taxonomy. */
public void testMixedRangeAndNonRangeTaxonomy() throws Exception {
    Directory d = newDirectory();
    RandomIndexWriter w = new RandomIndexWriter(random(), d);
    Directory td = newDirectory();
    DirectoryTaxonomyWriter tw = new DirectoryTaxonomyWriter(td, IndexWriterConfig.OpenMode.CREATE);
    FacetsConfig config = new FacetsConfig();
    for (long l = 0; l < 100; l++) {
        Document doc = new Document();
        // For computing range facet counts:
        doc.add(new NumericDocValuesField("field", l));
        // For drill down by numeric range:
        doc.add(new LongPoint("field", l));
        if ((l & 3) == 0) {
            doc.add(new FacetField("dim", "a"));
        } else {
            doc.add(new FacetField("dim", "b"));
        }
        w.addDocument(config.build(tw, doc));
    }
    final IndexReader r = w.getReader();
    final TaxonomyReader tr = new DirectoryTaxonomyReader(tw);
    IndexSearcher s = newSearcher(r, false);
    if (VERBOSE) {
        System.out.println("TEST: searcher=" + s);
    }
    DrillSideways ds = new DrillSideways(s, config, tr) {

        @Override
        protected Facets buildFacetsResult(FacetsCollector drillDowns, FacetsCollector[] drillSideways, String[] drillSidewaysDims) throws IOException {
            FacetsCollector dimFC = drillDowns;
            FacetsCollector fieldFC = drillDowns;
            if (drillSideways != null) {
                for (int i = 0; i < drillSideways.length; i++) {
                    String dim = drillSidewaysDims[i];
                    if (dim.equals("field")) {
                        fieldFC = drillSideways[i];
                    } else {
                        dimFC = drillSideways[i];
                    }
                }
            }
            Map<String, Facets> byDim = new HashMap<>();
            byDim.put("field", new LongRangeFacetCounts("field", fieldFC, new LongRange("less than 10", 0L, true, 10L, false), new LongRange("less than or equal to 10", 0L, true, 10L, true), new LongRange("over 90", 90L, false, 100L, false), new LongRange("90 or above", 90L, true, 100L, false), new LongRange("over 1000", 1000L, false, Long.MAX_VALUE, false)));
            byDim.put("dim", getTaxonomyFacetCounts(taxoReader, config, dimFC));
            return new MultiFacets(byDim, null);
        }

        @Override
        protected boolean scoreSubDocsAtOnce() {
            return random().nextBoolean();
        }
    };
    // First search, no drill downs:
    DrillDownQuery ddq = new DrillDownQuery(config);
    DrillSidewaysResult dsr = ds.search(null, ddq, 10);
    assertEquals(100, dsr.hits.totalHits);
    assertEquals("dim=dim path=[] value=100 childCount=2\n  b (75)\n  a (25)\n", dsr.facets.getTopChildren(10, "dim").toString());
    assertEquals("dim=field path=[] value=21 childCount=5\n  less than 10 (10)\n  less than or equal to 10 (11)\n  over 90 (9)\n  90 or above (10)\n  over 1000 (0)\n", dsr.facets.getTopChildren(10, "field").toString());
    // Second search, drill down on dim=b:
    ddq = new DrillDownQuery(config);
    ddq.add("dim", "b");
    dsr = ds.search(null, ddq, 10);
    assertEquals(75, dsr.hits.totalHits);
    assertEquals("dim=dim path=[] value=100 childCount=2\n  b (75)\n  a (25)\n", dsr.facets.getTopChildren(10, "dim").toString());
    assertEquals("dim=field path=[] value=16 childCount=5\n  less than 10 (7)\n  less than or equal to 10 (8)\n  over 90 (7)\n  90 or above (8)\n  over 1000 (0)\n", dsr.facets.getTopChildren(10, "field").toString());
    // Third search, drill down on "less than or equal to 10":
    ddq = new DrillDownQuery(config);
    ddq.add("field", LongPoint.newRangeQuery("field", 0L, 10L));
    dsr = ds.search(null, ddq, 10);
    assertEquals(11, dsr.hits.totalHits);
    assertEquals("dim=dim path=[] value=11 childCount=2\n  b (8)\n  a (3)\n", dsr.facets.getTopChildren(10, "dim").toString());
    assertEquals("dim=field path=[] value=21 childCount=5\n  less than 10 (10)\n  less than or equal to 10 (11)\n  over 90 (9)\n  90 or above (10)\n  over 1000 (0)\n", dsr.facets.getTopChildren(10, "field").toString());
    w.close();
    IOUtils.close(tw, tr, td, r, d);
}

Example 35 with FacetsConfig

use of org.apache.lucene.facet.FacetsConfig in project lucene-solr by apache.

the class TestSortedSetDocValuesFacets method testStaleState.

// LUCENE-5090
@SuppressWarnings("unused")
public void testStaleState() throws Exception {
    Directory dir = newDirectory();
    RandomIndexWriter writer = new RandomIndexWriter(random(), dir);
    FacetsConfig config = new FacetsConfig();
    Document doc = new Document();
    doc.add(new SortedSetDocValuesFacetField("a", "foo"));
    writer.addDocument(config.build(doc));
    IndexReader r = writer.getReader();
    SortedSetDocValuesReaderState state = new DefaultSortedSetDocValuesReaderState(r);
    doc = new Document();
    doc.add(new SortedSetDocValuesFacetField("a", "bar"));
    writer.addDocument(config.build(doc));
    doc = new Document();
    doc.add(new SortedSetDocValuesFacetField("a", "baz"));
    writer.addDocument(config.build(doc));
    IndexSearcher searcher = newSearcher(writer.getReader());
    FacetsCollector c = new FacetsCollector();
    searcher.search(new MatchAllDocsQuery(), c);
    expectThrows(IllegalStateException.class, () -> {
        new SortedSetDocValuesFacetCounts(state, c);
    });
    r.close();
    writer.close();
    searcher.getIndexReader().close();
    dir.close();
}