use of org.apache.lucene.facet.FacetResult in project lucene-solr by apache.
the class TestTaxonomyFacetCounts method testSparseFacets.
// LUCENE-5333
public void testSparseFacets() 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);
RandomIndexWriter writer = new RandomIndexWriter(random(), dir);
FacetsConfig config = new FacetsConfig();
Document doc = new Document();
doc.add(new FacetField("a", "foo1"));
writer.addDocument(config.build(taxoWriter, doc));
if (random().nextBoolean()) {
writer.commit();
}
doc = new Document();
doc.add(new FacetField("a", "foo2"));
doc.add(new FacetField("b", "bar1"));
writer.addDocument(config.build(taxoWriter, doc));
if (random().nextBoolean()) {
writer.commit();
}
doc = new Document();
doc.add(new FacetField("a", "foo3"));
doc.add(new FacetField("b", "bar2"));
doc.add(new FacetField("c", "baz1"));
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);
// 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());
writer.close();
IOUtils.close(taxoWriter, searcher.getIndexReader(), taxoReader, taxoDir, dir);
}
use of org.apache.lucene.facet.FacetResult in project lucene-solr by apache.
the class TestTaxonomyFacetCounts method testSegmentsWithoutCategoriesOrResults.
public void testSegmentsWithoutCategoriesOrResults() throws Exception {
// tests the accumulator when there are segments with no results
Directory indexDir = newDirectory();
Directory taxoDir = newDirectory();
IndexWriterConfig iwc = newIndexWriterConfig(new MockAnalyzer(random()));
// prevent merges
iwc.setMergePolicy(NoMergePolicy.INSTANCE);
IndexWriter indexWriter = new IndexWriter(indexDir, iwc);
TaxonomyWriter taxoWriter = new DirectoryTaxonomyWriter(taxoDir);
FacetsConfig config = new FacetsConfig();
// 1st segment, no content, with categories
indexTwoDocs(taxoWriter, indexWriter, config, false);
// 2nd segment, with content, no categories
indexTwoDocs(taxoWriter, indexWriter, null, true);
// 3rd segment ok
indexTwoDocs(taxoWriter, indexWriter, config, true);
// 4th segment, no content, or categories
indexTwoDocs(taxoWriter, indexWriter, null, false);
// 5th segment, with content, no categories
indexTwoDocs(taxoWriter, indexWriter, null, true);
// 6th segment, with content, with categories
indexTwoDocs(taxoWriter, indexWriter, config, true);
// 7th segment, with content, no categories
indexTwoDocs(taxoWriter, indexWriter, null, true);
indexWriter.close();
IOUtils.close(taxoWriter);
DirectoryReader indexReader = DirectoryReader.open(indexDir);
TaxonomyReader taxoReader = new DirectoryTaxonomyReader(taxoDir);
IndexSearcher indexSearcher = newSearcher(indexReader);
// search for "f:a", only segments 1 and 3 should match results
Query q = new TermQuery(new Term("f", "a"));
FacetsCollector sfc = new FacetsCollector();
indexSearcher.search(q, sfc);
Facets facets = getTaxonomyFacetCounts(taxoReader, config, sfc);
FacetResult result = facets.getTopChildren(10, "A");
assertEquals("wrong number of children", 2, result.labelValues.length);
for (LabelAndValue labelValue : result.labelValues) {
assertEquals("wrong weight for child " + labelValue.label, 2, labelValue.value.intValue());
}
IOUtils.close(indexReader, taxoReader, indexDir, taxoDir);
}
use of org.apache.lucene.facet.FacetResult in project lucene-solr by apache.
the class TestTaxonomyFacetCounts method testGetFacetResultsTwice.
public void testGetFacetResultsTwice() throws Exception {
// LUCENE-4893: counts were multiplied as many times as getFacetResults was called.
Directory indexDir = newDirectory();
Directory taxoDir = newDirectory();
DirectoryTaxonomyWriter taxoWriter = new DirectoryTaxonomyWriter(taxoDir);
IndexWriter iw = new IndexWriter(indexDir, newIndexWriterConfig(new MockAnalyzer(random())));
FacetsConfig config = new FacetsConfig();
Document doc = new Document();
doc.add(new FacetField("a", "1"));
doc.add(new FacetField("b", "1"));
iw.addDocument(config.build(taxoWriter, doc));
DirectoryReader r = DirectoryReader.open(iw);
DirectoryTaxonomyReader taxoReader = new DirectoryTaxonomyReader(taxoWriter);
Facets facets = getAllFacets(FacetsConfig.DEFAULT_INDEX_FIELD_NAME, newSearcher(r), taxoReader, config);
List<FacetResult> res1 = facets.getAllDims(10);
List<FacetResult> res2 = facets.getAllDims(10);
assertEquals("calling getFacetResults twice should return the .equals()=true result", res1, res2);
iw.close();
IOUtils.close(taxoWriter, taxoReader, taxoDir, r, indexDir);
}
use of org.apache.lucene.facet.FacetResult in project lucene-solr by apache.
the class TestTaxonomyFacetCounts method testCountRoot.
public void testCountRoot() throws Exception {
// LUCENE-4882: FacetsAccumulator threw NPE if a FacetRequest was defined on CP.EMPTY
Directory indexDir = newDirectory();
Directory taxoDir = newDirectory();
DirectoryTaxonomyWriter taxoWriter = new DirectoryTaxonomyWriter(taxoDir);
IndexWriter iw = new IndexWriter(indexDir, newIndexWriterConfig(new MockAnalyzer(random())));
FacetsConfig config = new FacetsConfig();
for (int i = atLeast(30); i > 0; --i) {
Document doc = new Document();
doc.add(new FacetField("a", "1"));
doc.add(new FacetField("b", "1"));
iw.addDocument(config.build(taxoWriter, doc));
}
DirectoryReader r = DirectoryReader.open(iw);
DirectoryTaxonomyReader taxoReader = new DirectoryTaxonomyReader(taxoWriter);
Facets facets = getAllFacets(FacetsConfig.DEFAULT_INDEX_FIELD_NAME, newSearcher(r), taxoReader, config);
for (FacetResult result : facets.getAllDims(10)) {
assertEquals(r.numDocs(), result.value.intValue());
}
iw.close();
IOUtils.close(taxoWriter, taxoReader, taxoDir, r, indexDir);
}
use of org.apache.lucene.facet.FacetResult in project lucene-solr by apache.
the class TestTaxonomyFacetCounts method testManyFacetsInOneDocument.
// LUCENE-4583: make sure if we require > 32 KB for one
// document, we don't hit exc when using Facet42DocValuesFormat
public void testManyFacetsInOneDocument() throws Exception {
assumeTrue("default Codec doesn't support huge BinaryDocValues", TestUtil.fieldSupportsHugeBinaryDocValues(FacetsConfig.DEFAULT_INDEX_FIELD_NAME));
Directory dir = newDirectory();
Directory taxoDir = newDirectory();
IndexWriterConfig iwc = newIndexWriterConfig(new MockAnalyzer(random()));
RandomIndexWriter writer = new RandomIndexWriter(random(), dir, iwc);
DirectoryTaxonomyWriter taxoWriter = new DirectoryTaxonomyWriter(taxoDir, IndexWriterConfig.OpenMode.CREATE);
FacetsConfig config = new FacetsConfig();
config.setMultiValued("dim", true);
int numLabels = TestUtil.nextInt(random(), 40000, 100000);
Document doc = new Document();
doc.add(newTextField("field", "text", Field.Store.NO));
for (int i = 0; i < numLabels; i++) {
doc.add(new FacetField("dim", "" + i));
}
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);
FacetResult result = facets.getTopChildren(Integer.MAX_VALUE, "dim");
assertEquals(numLabels, result.labelValues.length);
Set<String> allLabels = new HashSet<>();
for (LabelAndValue labelValue : result.labelValues) {
allLabels.add(labelValue.label);
assertEquals(1, labelValue.value.intValue());
}
assertEquals(numLabels, allLabels.size());
writer.close();
IOUtils.close(searcher.getIndexReader(), taxoWriter, taxoReader, dir, taxoDir);
}
Aggregations