1 package org.apache.lucene.search.payloads;
3 * Licensed to the Apache Software Foundation (ASF) under one or more
4 * contributor license agreements. See the NOTICE file distributed with
5 * this work for additional information regarding copyright ownership.
6 * The ASF licenses this file to You under the Apache License, Version 2.0
7 * (the "License"); you may not use this file except in compliance with
8 * the License. You may obtain a copy of the License at
10 * http://www.apache.org/licenses/LICENSE-2.0
12 * Unless required by applicable law or agreed to in writing, software
13 * distributed under the License is distributed on an "AS IS" BASIS,
14 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
15 * See the License for the specific language governing permissions and
16 * limitations under the License.
18 import java.io.IOException;
19 import java.io.Reader;
20 import java.util.Collection;
22 import org.apache.lucene.analysis.Analyzer;
23 import org.apache.lucene.analysis.LowerCaseTokenizer;
24 import org.apache.lucene.analysis.TokenFilter;
25 import org.apache.lucene.analysis.TokenStream;
26 import org.apache.lucene.analysis.tokenattributes.PayloadAttribute;
27 import org.apache.lucene.document.Document;
28 import org.apache.lucene.document.Field;
29 import org.apache.lucene.index.FieldInvertState;
30 import org.apache.lucene.index.IndexReader;
31 import org.apache.lucene.index.Payload;
32 import org.apache.lucene.index.RandomIndexWriter;
33 import org.apache.lucene.index.Term;
34 import org.apache.lucene.search.DefaultSimilarity;
35 import org.apache.lucene.search.Explanation;
36 import org.apache.lucene.search.IndexSearcher;
37 import org.apache.lucene.search.QueryUtils;
38 import org.apache.lucene.search.ScoreDoc;
39 import org.apache.lucene.search.Searcher;
40 import org.apache.lucene.search.TopDocs;
41 import org.apache.lucene.search.spans.SpanQuery;
42 import org.apache.lucene.search.spans.SpanNearQuery;
43 import org.apache.lucene.search.spans.SpanTermQuery;
44 import org.apache.lucene.store.Directory;
45 import org.apache.lucene.util.English;
46 import org.apache.lucene.util.LuceneTestCase;
47 import org.apache.lucene.search.Explanation.IDFExplanation;
48 import org.junit.AfterClass;
49 import org.junit.BeforeClass;
52 public class TestPayloadNearQuery extends LuceneTestCase {
53 private static IndexSearcher searcher;
54 private static IndexReader reader;
55 private static Directory directory;
56 private static BoostingSimilarity similarity = new BoostingSimilarity();
57 private static byte[] payload2 = new byte[]{2};
58 private static byte[] payload4 = new byte[]{4};
60 private static class PayloadAnalyzer extends Analyzer {
62 public TokenStream tokenStream(String fieldName, Reader reader) {
63 TokenStream result = new LowerCaseTokenizer(TEST_VERSION_CURRENT, reader);
64 result = new PayloadFilter(result, fieldName);
69 private static class PayloadFilter extends TokenFilter {
72 protected PayloadAttribute payAtt;
74 public PayloadFilter(TokenStream input, String fieldName) {
76 this.fieldName = fieldName;
77 payAtt = addAttribute(PayloadAttribute.class);
81 public boolean incrementToken() throws IOException {
82 boolean result = false;
83 if (input.incrementToken() == true){
84 if (numSeen % 2 == 0) {
85 payAtt.setPayload(new Payload(payload2));
87 payAtt.setPayload(new Payload(payload4));
96 private PayloadNearQuery newPhraseQuery (String fieldName, String phrase, boolean inOrder, PayloadFunction function ) {
97 String[] words = phrase.split("[\\s]+");
98 SpanQuery clauses[] = new SpanQuery[words.length];
99 for (int i=0;i<clauses.length;i++) {
100 clauses[i] = new SpanTermQuery(new Term(fieldName, words[i]));
102 return new PayloadNearQuery(clauses, 0, inOrder, function);
106 public static void beforeClass() throws Exception {
107 directory = newDirectory();
108 RandomIndexWriter writer = new RandomIndexWriter(random, directory,
109 newIndexWriterConfig(TEST_VERSION_CURRENT, new PayloadAnalyzer())
110 .setSimilarity(similarity));
111 //writer.infoStream = System.out;
112 for (int i = 0; i < 1000; i++) {
113 Document doc = new Document();
114 doc.add(newField("field", English.intToEnglish(i), Field.Store.YES, Field.Index.ANALYZED));
115 String txt = English.intToEnglish(i) +' '+English.intToEnglish(i+1);
116 doc.add(newField("field2", txt, Field.Store.YES, Field.Index.ANALYZED));
117 writer.addDocument(doc);
119 reader = writer.getReader();
122 searcher = newSearcher(reader);
123 searcher.setSimilarity(similarity);
127 public static void afterClass() throws Exception {
136 public void test() throws IOException {
137 PayloadNearQuery query;
140 query = newPhraseQuery("field", "twenty two", true, new AveragePayloadFunction());
141 QueryUtils.check(query);
143 // all 10 hits should have score = 3 because adjacent terms have payloads of 2,4
144 // and all the similarity factors are set to 1
145 hits = searcher.search(query, null, 100);
146 assertTrue("hits is null and it shouldn't be", hits != null);
147 assertTrue("should be 10 hits", hits.totalHits == 10);
148 for (int j = 0; j < hits.scoreDocs.length; j++) {
149 ScoreDoc doc = hits.scoreDocs[j];
150 assertTrue(doc.score + " does not equal: " + 3, doc.score == 3);
152 for (int i=1;i<10;i++) {
153 query = newPhraseQuery("field", English.intToEnglish(i)+" hundred", true, new AveragePayloadFunction());
154 // all should have score = 3 because adjacent terms have payloads of 2,4
155 // and all the similarity factors are set to 1
156 hits = searcher.search(query, null, 100);
157 assertTrue("hits is null and it shouldn't be", hits != null);
158 assertTrue("should be 100 hits", hits.totalHits == 100);
159 for (int j = 0; j < hits.scoreDocs.length; j++) {
160 ScoreDoc doc = hits.scoreDocs[j];
161 // System.out.println("Doc: " + doc.toString());
162 // System.out.println("Explain: " + searcher.explain(query, doc.doc));
163 assertTrue(doc.score + " does not equal: " + 3, doc.score == 3);
169 public void testPayloadNear() throws IOException {
170 SpanNearQuery q1, q2;
171 PayloadNearQuery query;
172 //SpanNearQuery(clauses, 10000, false)
173 q1 = spanNearQuery("field2", "twenty two");
174 q2 = spanNearQuery("field2", "twenty three");
175 SpanQuery[] clauses = new SpanQuery[2];
178 query = new PayloadNearQuery(clauses, 10, false);
179 //System.out.println(query.toString());
180 assertEquals(12, searcher.search(query, null, 100).totalHits);
182 System.out.println(hits.totalHits);
183 for (int j = 0; j < hits.scoreDocs.length; j++) {
184 ScoreDoc doc = hits.scoreDocs[j];
185 System.out.println("doc: "+doc.doc+", score: "+doc.score);
190 public void testAverageFunction() throws IOException {
191 PayloadNearQuery query;
194 query = newPhraseQuery("field", "twenty two", true, new AveragePayloadFunction());
195 QueryUtils.check(query);
196 // all 10 hits should have score = 3 because adjacent terms have payloads of 2,4
197 // and all the similarity factors are set to 1
198 hits = searcher.search(query, null, 100);
199 assertTrue("hits is null and it shouldn't be", hits != null);
200 assertTrue("should be 10 hits", hits.totalHits == 10);
201 for (int j = 0; j < hits.scoreDocs.length; j++) {
202 ScoreDoc doc = hits.scoreDocs[j];
203 assertTrue(doc.score + " does not equal: " + 3, doc.score == 3);
204 Explanation explain = searcher.explain(query, hits.scoreDocs[j].doc);
205 String exp = explain.toString();
206 assertTrue(exp, exp.indexOf("AveragePayloadFunction") > -1);
207 assertTrue(hits.scoreDocs[j].score + " explain value does not equal: " + 3, explain.getValue() == 3f);
210 public void testMaxFunction() throws IOException {
211 PayloadNearQuery query;
214 query = newPhraseQuery("field", "twenty two", true, new MaxPayloadFunction());
215 QueryUtils.check(query);
216 // all 10 hits should have score = 4 (max payload value)
217 hits = searcher.search(query, null, 100);
218 assertTrue("hits is null and it shouldn't be", hits != null);
219 assertTrue("should be 10 hits", hits.totalHits == 10);
220 for (int j = 0; j < hits.scoreDocs.length; j++) {
221 ScoreDoc doc = hits.scoreDocs[j];
222 assertTrue(doc.score + " does not equal: " + 4, doc.score == 4);
223 Explanation explain = searcher.explain(query, hits.scoreDocs[j].doc);
224 String exp = explain.toString();
225 assertTrue(exp, exp.indexOf("MaxPayloadFunction") > -1);
226 assertTrue(hits.scoreDocs[j].score + " explain value does not equal: " + 4, explain.getValue() == 4f);
229 public void testMinFunction() throws IOException {
230 PayloadNearQuery query;
233 query = newPhraseQuery("field", "twenty two", true, new MinPayloadFunction());
234 QueryUtils.check(query);
235 // all 10 hits should have score = 2 (min payload value)
236 hits = searcher.search(query, null, 100);
237 assertTrue("hits is null and it shouldn't be", hits != null);
238 assertTrue("should be 10 hits", hits.totalHits == 10);
239 for (int j = 0; j < hits.scoreDocs.length; j++) {
240 ScoreDoc doc = hits.scoreDocs[j];
241 assertTrue(doc.score + " does not equal: " + 2, doc.score == 2);
242 Explanation explain = searcher.explain(query, hits.scoreDocs[j].doc);
243 String exp = explain.toString();
244 assertTrue(exp, exp.indexOf("MinPayloadFunction") > -1);
245 assertTrue(hits.scoreDocs[j].score + " explain value does not equal: " + 2, explain.getValue() == 2f);
248 private SpanQuery[] getClauses() {
249 SpanNearQuery q1, q2;
250 q1 = spanNearQuery("field2", "twenty two");
251 q2 = spanNearQuery("field2", "twenty three");
252 SpanQuery[] clauses = new SpanQuery[2];
257 private SpanNearQuery spanNearQuery(String fieldName, String words) {
258 String[] wordList = words.split("[\\s]+");
259 SpanQuery clauses[] = new SpanQuery[wordList.length];
260 for (int i=0;i<clauses.length;i++) {
261 clauses[i] = new PayloadTermQuery(new Term(fieldName, wordList[i]), new AveragePayloadFunction());
263 return new SpanNearQuery(clauses, 10000, false);
266 public void testLongerSpan() throws IOException {
267 PayloadNearQuery query;
269 query = newPhraseQuery("field", "nine hundred ninety nine", true, new AveragePayloadFunction());
270 hits = searcher.search(query, null, 100);
271 assertTrue("hits is null and it shouldn't be", hits != null);
272 ScoreDoc doc = hits.scoreDocs[0];
273 // System.out.println("Doc: " + doc.toString());
274 // System.out.println("Explain: " + searcher.explain(query, doc.doc));
275 assertTrue("there should only be one hit", hits.totalHits == 1);
276 // should have score = 3 because adjacent terms have payloads of 2,4
277 assertTrue(doc.score + " does not equal: " + 3, doc.score == 3);
280 public void testComplexNested() throws IOException {
281 PayloadNearQuery query;
284 // combine ordered and unordered spans with some nesting to make sure all payloads are counted
286 SpanQuery q1 = newPhraseQuery("field", "nine hundred", true, new AveragePayloadFunction());
287 SpanQuery q2 = newPhraseQuery("field", "ninety nine", true, new AveragePayloadFunction());
288 SpanQuery q3 = newPhraseQuery("field", "nine ninety", false, new AveragePayloadFunction());
289 SpanQuery q4 = newPhraseQuery("field", "hundred nine", false, new AveragePayloadFunction());
290 SpanQuery[]clauses = new SpanQuery[] {new PayloadNearQuery(new SpanQuery[] {q1,q2}, 0, true), new PayloadNearQuery(new SpanQuery[] {q3,q4}, 0, false)};
291 query = new PayloadNearQuery(clauses, 0, false);
292 hits = searcher.search(query, null, 100);
293 assertTrue("hits is null and it shouldn't be", hits != null);
294 // should be only 1 hit - doc 999
295 assertTrue("should only be one hit", hits.scoreDocs.length == 1);
296 // the score should be 3 - the average of all the underlying payloads
297 ScoreDoc doc = hits.scoreDocs[0];
298 // System.out.println("Doc: " + doc.toString());
299 // System.out.println("Explain: " + searcher.explain(query, doc.doc));
300 assertTrue(doc.score + " does not equal: " + 3, doc.score == 3);
303 // must be static for weight serialization tests
304 static class BoostingSimilarity extends DefaultSimilarity {
306 @Override public float scorePayload(int docId, String fieldName, int start, int end, byte[] payload, int offset, int length) {
307 //we know it is size 4 here, so ignore the offset/length
310 //!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
311 //Make everything else 1 so we see the effect of the payload
312 //!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
313 @Override public float computeNorm(String fieldName, FieldInvertState state) {
314 return state.getBoost();
317 @Override public float queryNorm(float sumOfSquaredWeights) {
321 @Override public float sloppyFreq(int distance) {
325 @Override public float coord(int overlap, int maxOverlap) {
328 @Override public float tf(float freq) {
331 // idf used for phrase queries
332 @Override public IDFExplanation idfExplain(Collection<Term> terms, Searcher searcher) throws IOException {
333 return new IDFExplanation() {
335 public float getIdf() {
339 public String explain() {
340 return "Inexplicable";