/* * Licensed to the Apache Software Foundation (ASF) under one or more * contributor license agreements. See the NOTICE file distributed with * this work for additional information regarding copyright ownership. * The ASF licenses this file to You under the Apache License, Version 2.0 * (the "License"); you may not use this file except in compliance with * the License. You may obtain a copy of the License at * * http://www.apache.org/licenses/LICENSE-2.0 * * Unless required by applicable law or agreed to in writing, software * distributed under the License is distributed on an "AS IS" BASIS, * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. * See the License for the specific language governing permissions and * limitations under the License. */ package org.apache.lucene.search.grouping; import java.util.HashMap; import org.apache.lucene.analysis.MockAnalyzer; import org.apache.lucene.document.Document; import org.apache.lucene.document.Field; import org.apache.lucene.document.FieldType; import org.apache.lucene.document.SortedDocValuesField; import org.apache.lucene.document.TextField; import org.apache.lucene.index.RandomIndexWriter; import org.apache.lucene.index.Term; import org.apache.lucene.queries.function.ValueSource; import org.apache.lucene.queries.function.valuesource.BytesRefFieldSource; import org.apache.lucene.search.IndexSearcher; import org.apache.lucene.search.TermQuery; import org.apache.lucene.store.Directory; import org.apache.lucene.util.BytesRef; import org.apache.lucene.util.LuceneTestCase; public class AllGroupsCollectorTest extends LuceneTestCase { public void testTotalGroupCount() throws Exception { final String groupField = "author"; FieldType customType = new FieldType(); customType.setStored(true); Directory dir = newDirectory(); RandomIndexWriter w = new RandomIndexWriter( random(), dir, newIndexWriterConfig(new MockAnalyzer(random())).setMergePolicy(newLogMergePolicy())); // 0 Document doc = new Document(); addGroupField(doc, groupField, "author1"); doc.add(new TextField("content", "random text", Field.Store.YES)); doc.add(new Field("id", "1", customType)); w.addDocument(doc); // 1 doc = new Document(); addGroupField(doc, groupField, "author1"); doc.add(new TextField("content", "some more random text blob", Field.Store.YES)); doc.add(new Field("id", "2", customType)); w.addDocument(doc); // 2 doc = new Document(); addGroupField(doc, groupField, "author1"); doc.add(new TextField("content", "some more random textual data", Field.Store.YES)); doc.add(new Field("id", "3", customType)); w.addDocument(doc); w.commit(); // To ensure a second segment // 3 doc = new Document(); addGroupField(doc, groupField, "author2"); doc.add(new TextField("content", "some random text", Field.Store.YES)); doc.add(new Field("id", "4", customType)); w.addDocument(doc); // 4 doc = new Document(); addGroupField(doc, groupField, "author3"); doc.add(new TextField("content", "some more random text", Field.Store.YES)); doc.add(new Field("id", "5", customType)); w.addDocument(doc); // 5 doc = new Document(); addGroupField(doc, groupField, "author3"); doc.add(new TextField("content", "random blob", Field.Store.YES)); doc.add(new Field("id", "6", customType)); w.addDocument(doc); // 6 -- no author field doc = new Document(); doc.add(new TextField("content", "random word stuck in alot of other text", Field.Store.YES)); doc.add(new Field("id", "6", customType)); w.addDocument(doc); IndexSearcher indexSearcher = newSearcher(w.getReader()); w.close(); AllGroupsCollector<?> allGroupsCollector = createRandomCollector(groupField); indexSearcher.search(new TermQuery(new Term("content", "random")), allGroupsCollector); assertEquals(4, allGroupsCollector.getGroupCount()); allGroupsCollector = createRandomCollector(groupField); indexSearcher.search(new TermQuery(new Term("content", "some")), allGroupsCollector); assertEquals(3, allGroupsCollector.getGroupCount()); allGroupsCollector = createRandomCollector(groupField); indexSearcher.search(new TermQuery(new Term("content", "blob")), allGroupsCollector); assertEquals(2, allGroupsCollector.getGroupCount()); indexSearcher.getIndexReader().close(); dir.close(); } private void addGroupField(Document doc, String groupField, String value) { doc.add(new TextField(groupField, value, Field.Store.YES)); doc.add(new SortedDocValuesField(groupField, new BytesRef(value))); } private AllGroupsCollector<?> createRandomCollector(String groupField) { if (random().nextBoolean()) { return new AllGroupsCollector<>(new TermGroupSelector(groupField)); } else { ValueSource vs = new BytesRefFieldSource(groupField); return new AllGroupsCollector<>(new ValueSourceGroupSelector(vs, new HashMap<>())); } } }