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

Example 11 with FacetResult

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

the class TestSortedSetDocValuesFacets method testRandom.

public void testRandom() throws Exception {
    String[] tokens = getRandomTokens(10);
    Directory indexDir = newDirectory();
    Directory taxoDir = newDirectory();
    RandomIndexWriter w = new RandomIndexWriter(random(), indexDir);
    FacetsConfig config = new FacetsConfig();
    int numDocs = atLeast(1000);
    int numDims = TestUtil.nextInt(random(), 1, 7);
    List<TestDoc> testDocs = getRandomDocs(tokens, numDocs, numDims);
    for (TestDoc testDoc : testDocs) {
        Document doc = new Document();
        doc.add(newStringField("content", testDoc.content, Field.Store.NO));
        for (int j = 0; j < numDims; j++) {
            if (testDoc.dims[j] != null) {
                doc.add(new SortedSetDocValuesFacetField("dim" + j, testDoc.dims[j]));
            }
        }
        w.addDocument(config.build(doc));
    }
    // NRT open
    IndexSearcher searcher = newSearcher(w.getReader());
    // Per-top-reader state:
    SortedSetDocValuesReaderState state = new DefaultSortedSetDocValuesReaderState(searcher.getIndexReader());
    ExecutorService exec = randomExecutorServiceOrNull();
    int iters = atLeast(100);
    for (int iter = 0; iter < iters; iter++) {
        String searchToken = tokens[random().nextInt(tokens.length)];
        if (VERBOSE) {
            System.out.println("\nTEST: iter content=" + searchToken);
        }
        FacetsCollector fc = new FacetsCollector();
        FacetsCollector.search(searcher, new TermQuery(new Term("content", searchToken)), 10, fc);
        Facets facets;
        if (exec != null) {
            facets = new ConcurrentSortedSetDocValuesFacetCounts(state, fc, exec);
        } else {
            facets = new SortedSetDocValuesFacetCounts(state, fc);
        }
        // Slow, yet hopefully bug-free, faceting:
        @SuppressWarnings({ "rawtypes", "unchecked" }) Map<String, Integer>[] expectedCounts = new HashMap[numDims];
        for (int i = 0; i < numDims; i++) {
            expectedCounts[i] = new HashMap<>();
        }
        for (TestDoc doc : testDocs) {
            if (doc.content.equals(searchToken)) {
                for (int j = 0; j < numDims; j++) {
                    if (doc.dims[j] != null) {
                        Integer v = expectedCounts[j].get(doc.dims[j]);
                        if (v == null) {
                            expectedCounts[j].put(doc.dims[j], 1);
                        } else {
                            expectedCounts[j].put(doc.dims[j], v.intValue() + 1);
                        }
                    }
                }
            }
        }
        List<FacetResult> expected = new ArrayList<>();
        for (int i = 0; i < numDims; i++) {
            List<LabelAndValue> labelValues = new ArrayList<>();
            int totCount = 0;
            for (Map.Entry<String, Integer> ent : expectedCounts[i].entrySet()) {
                labelValues.add(new LabelAndValue(ent.getKey(), ent.getValue()));
                totCount += ent.getValue();
            }
            sortLabelValues(labelValues);
            if (totCount > 0) {
                expected.add(new FacetResult("dim" + i, new String[0], totCount, labelValues.toArray(new LabelAndValue[labelValues.size()]), labelValues.size()));
            }
        }
        // Sort by highest value, tie break by value:
        sortFacetResults(expected);
        List<FacetResult> actual = facets.getAllDims(10);
        // Messy: fixup ties
        //sortTies(actual);
        assertEquals(expected, actual);
    }
    if (exec != null) {
        exec.shutdownNow();
    }
    w.close();
    IOUtils.close(searcher.getIndexReader(), indexDir, taxoDir);
}

Example 12 with FacetResult

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

the class TestSortedSetDocValuesFacets method testSparseFacets.

// LUCENE-5333
public void testSparseFacets() throws Exception {
    Directory dir = newDirectory();
    RandomIndexWriter writer = new RandomIndexWriter(random(), dir);
    FacetsConfig config = new FacetsConfig();
    Document doc = new Document();
    doc.add(new SortedSetDocValuesFacetField("a", "foo1"));
    writer.addDocument(config.build(doc));
    if (random().nextBoolean()) {
        writer.commit();
    }
    doc = new Document();
    doc.add(new SortedSetDocValuesFacetField("a", "foo2"));
    doc.add(new SortedSetDocValuesFacetField("b", "bar1"));
    writer.addDocument(config.build(doc));
    if (random().nextBoolean()) {
        writer.commit();
    }
    doc = new Document();
    doc.add(new SortedSetDocValuesFacetField("a", "foo3"));
    doc.add(new SortedSetDocValuesFacetField("b", "bar2"));
    doc.add(new SortedSetDocValuesFacetField("c", "baz1"));
    writer.addDocument(config.build(doc));
    // NRT open
    IndexSearcher searcher = newSearcher(writer.getReader());
    writer.close();
    // Per-top-reader state:
    SortedSetDocValuesReaderState state = new DefaultSortedSetDocValuesReaderState(searcher.getIndexReader());
    ExecutorService exec = randomExecutorServiceOrNull();
    Facets facets = getAllFacets(searcher, state, exec);
    // Ask for top 10 labels for any dims that have counts:
    List<FacetResult> results = facets.getAllDims(10);
    assertEquals(3, results.size());
    assertEquals("dim=a path=[] value=3 childCount=3\n  foo1 (1)\n  foo2 (1)\n  foo3 (1)\n", results.get(0).toString());
    assertEquals("dim=b path=[] value=2 childCount=2\n  bar1 (1)\n  bar2 (1)\n", results.get(1).toString());
    assertEquals("dim=c path=[] value=1 childCount=1\n  baz1 (1)\n", results.get(2).toString());
    Collection<Accountable> resources = state.getChildResources();
    assertTrue(state.toString().contains(FacetsConfig.DEFAULT_INDEX_FIELD_NAME));
    if (searcher.getIndexReader().leaves().size() > 1) {
        assertTrue(state.ramBytesUsed() > 0);
        assertFalse(resources.isEmpty());
        assertTrue(resources.toString().contains(FacetsConfig.DEFAULT_INDEX_FIELD_NAME));
    } else {
        assertEquals(0, state.ramBytesUsed());
        assertTrue(resources.isEmpty());
    }
    if (exec != null) {
        exec.shutdownNow();
    }
    searcher.getIndexReader().close();
    dir.close();
}

Example 13 with FacetResult

use of org.apache.lucene.facet.FacetResult 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);
}

Example 14 with FacetResult

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

the class TestTaxonomyFacetCounts method testRandom.

public void testRandom() throws Exception {
    String[] tokens = getRandomTokens(10);
    Directory indexDir = newDirectory();
    Directory taxoDir = newDirectory();
    RandomIndexWriter w = new RandomIndexWriter(random(), indexDir);
    DirectoryTaxonomyWriter tw = new DirectoryTaxonomyWriter(taxoDir);
    FacetsConfig config = new FacetsConfig();
    int numDocs = atLeast(1000);
    int numDims = TestUtil.nextInt(random(), 1, 7);
    List<TestDoc> testDocs = getRandomDocs(tokens, numDocs, numDims);
    for (TestDoc testDoc : testDocs) {
        Document doc = new Document();
        doc.add(newStringField("content", testDoc.content, Field.Store.NO));
        for (int j = 0; j < numDims; j++) {
            if (testDoc.dims[j] != null) {
                doc.add(new FacetField("dim" + j, testDoc.dims[j]));
            }
        }
        w.addDocument(config.build(tw, doc));
    }
    // NRT open
    IndexSearcher searcher = newSearcher(w.getReader());
    // NRT open
    TaxonomyReader tr = new DirectoryTaxonomyReader(tw);
    int iters = atLeast(100);
    for (int iter = 0; iter < iters; iter++) {
        String searchToken = tokens[random().nextInt(tokens.length)];
        if (VERBOSE) {
            System.out.println("\nTEST: iter content=" + searchToken);
        }
        FacetsCollector fc = new FacetsCollector();
        FacetsCollector.search(searcher, new TermQuery(new Term("content", searchToken)), 10, fc);
        Facets facets = getTaxonomyFacetCounts(tr, config, fc);
        // Slow, yet hopefully bug-free, faceting:
        @SuppressWarnings({ "rawtypes", "unchecked" }) Map<String, Integer>[] expectedCounts = new HashMap[numDims];
        for (int i = 0; i < numDims; i++) {
            expectedCounts[i] = new HashMap<>();
        }
        for (TestDoc doc : testDocs) {
            if (doc.content.equals(searchToken)) {
                for (int j = 0; j < numDims; j++) {
                    if (doc.dims[j] != null) {
                        Integer v = expectedCounts[j].get(doc.dims[j]);
                        if (v == null) {
                            expectedCounts[j].put(doc.dims[j], 1);
                        } else {
                            expectedCounts[j].put(doc.dims[j], v.intValue() + 1);
                        }
                    }
                }
            }
        }
        List<FacetResult> expected = new ArrayList<>();
        for (int i = 0; i < numDims; i++) {
            List<LabelAndValue> labelValues = new ArrayList<>();
            int totCount = 0;
            for (Map.Entry<String, Integer> ent : expectedCounts[i].entrySet()) {
                labelValues.add(new LabelAndValue(ent.getKey(), ent.getValue()));
                totCount += ent.getValue();
            }
            sortLabelValues(labelValues);
            if (totCount > 0) {
                expected.add(new FacetResult("dim" + i, new String[0], totCount, labelValues.toArray(new LabelAndValue[labelValues.size()]), labelValues.size()));
            }
        }
        // Sort by highest value, tie break by value:
        sortFacetResults(expected);
        List<FacetResult> actual = facets.getAllDims(10);
        // Messy: fixup ties
        sortTies(actual);
        assertEquals(expected, actual);
    }
    w.close();
    IOUtils.close(tw, searcher.getIndexReader(), tr, indexDir, taxoDir);
}

Example 15 with FacetResult

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

the class TestTaxonomyFacetCounts method testMultiValuedHierarchy.

public void testMultiValuedHierarchy() throws Exception {
    Directory dir = newDirectory();
    Directory taxoDir = newDirectory();
    DirectoryTaxonomyWriter taxoWriter = new DirectoryTaxonomyWriter(taxoDir, IndexWriterConfig.OpenMode.CREATE);
    FacetsConfig config = new FacetsConfig();
    config.setHierarchical("a", true);
    config.setMultiValued("a", true);
    RandomIndexWriter writer = new RandomIndexWriter(random(), dir);
    Document doc = new Document();
    doc.add(newTextField("field", "text", Field.Store.NO));
    doc.add(new FacetField("a", "path", "x"));
    doc.add(new FacetField("a", "path", "y"));
    writer.addDocument(config.build(taxoWriter, doc));
    // NRT open
    IndexSearcher searcher = newSearcher(writer.getReader());
    // NRT open
    TaxonomyReader taxoReader = new DirectoryTaxonomyReader(taxoWriter);
    Facets facets = getAllFacets(FacetsConfig.DEFAULT_INDEX_FIELD_NAME, searcher, taxoReader, config);
    expectThrows(IllegalArgumentException.class, () -> {
        facets.getSpecificValue("a");
    });
    FacetResult result = facets.getTopChildren(10, "a");
    assertEquals(1, result.labelValues.length);
    assertEquals(1, result.labelValues[0].value.intValue());
    writer.close();
    IOUtils.close(taxoWriter, searcher.getIndexReader(), taxoReader, dir, taxoDir);
}