1
package org.apache.lucene.search;
4
* Copyright 2005 The Apache Software Foundation
6
* Licensed under the Apache License, Version 2.0 (the "License");
7
* you may not use this file except in compliance with the License.
8
* 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.
19
import java.io.IOException;
20
import java.io.Serializable;
21
import java.util.Calendar;
22
import java.util.GregorianCalendar;
24
import java.util.Random;
25
import java.util.TreeMap;
27
import org.apache.lucene.document.DateTools;
28
import org.apache.lucene.document.Document;
29
import org.apache.lucene.document.Field;
30
import org.apache.lucene.index.IndexReader;
31
import org.apache.lucene.index.RandomIndexWriter;
32
import org.apache.lucene.index.Term;
33
import org.apache.lucene.store.Directory;
34
import org.apache.lucene.util.LuceneTestCase;
36
/** Unit test for sorting code. */
37
public class TestCustomSearcherSort extends LuceneTestCase implements Serializable {
39
private Directory index = null;
40
private IndexReader reader;
41
private Query query = null;
42
// reduced from 20000 to 2000 to speed up test...
43
private int INDEX_SIZE;
46
* Create index and query for test cases.
49
public void setUp() throws Exception {
51
INDEX_SIZE = atLeast(2000);
52
index = newDirectory();
53
RandomIndexWriter writer = new RandomIndexWriter(random, index);
54
RandomGen random = new RandomGen(this.random);
55
for (int i = 0; i < INDEX_SIZE; ++i) { // don't decrease; if to low the
56
// problem doesn't show up
57
Document doc = new Document();
58
if ((i % 5) != 0) { // some documents must not have an entry in the first
60
doc.add(newField("publicationDate_", random.getLuceneDate(),
61
Field.Store.YES, Field.Index.NOT_ANALYZED));
63
if ((i % 7) == 0) { // some documents to match the query (see below)
64
doc.add(newField("content", "test", Field.Store.YES,
65
Field.Index.ANALYZED));
67
// every document has a defined 'mandant' field
68
doc.add(newField("mandant", Integer.toString(i % 3), Field.Store.YES,
69
Field.Index.NOT_ANALYZED));
70
writer.addDocument(doc);
72
reader = writer.getReader();
74
query = new TermQuery(new Term("content", "test"));
78
public void tearDown() throws Exception {
85
* Run the test using two CustomSearcher instances.
87
public void testFieldSortCustomSearcher() throws Exception {
88
// log("Run testFieldSortCustomSearcher");
89
// define the sort criteria
90
Sort custSort = new Sort(
91
new SortField("publicationDate_", SortField.STRING),
92
SortField.FIELD_SCORE);
93
Searcher searcher = new CustomSearcher(reader, 2);
94
// search and check hits
95
matchHits(searcher, custSort);
99
* Run the test using one CustomSearcher wrapped by a MultiSearcher.
101
public void testFieldSortSingleSearcher() throws Exception {
102
// log("Run testFieldSortSingleSearcher");
103
// define the sort criteria
104
Sort custSort = new Sort(
105
new SortField("publicationDate_", SortField.STRING),
106
SortField.FIELD_SCORE);
107
Searcher searcher = new MultiSearcher(new Searcher[] {new CustomSearcher(
109
// search and check hits
110
matchHits(searcher, custSort);
114
* Run the test using two CustomSearcher instances.
116
public void testFieldSortMultiCustomSearcher() throws Exception {
117
// log("Run testFieldSortMultiCustomSearcher");
118
// define the sort criteria
119
Sort custSort = new Sort(
120
new SortField("publicationDate_", SortField.STRING),
121
SortField.FIELD_SCORE);
122
Searcher searcher = new MultiSearcher(new Searchable[] {
123
new CustomSearcher(reader, 0), new CustomSearcher(reader, 2)});
124
// search and check hits
125
matchHits(searcher, custSort);
128
// make sure the documents returned by the search match the expected list
129
private void matchHits(Searcher searcher, Sort sort) throws IOException {
130
// make a query without sorting first
131
ScoreDoc[] hitsByRank = searcher.search(query, null, Integer.MAX_VALUE).scoreDocs;
132
checkHits(hitsByRank, "Sort by rank: "); // check for duplicates
133
Map<Integer,Integer> resultMap = new TreeMap<Integer,Integer>();
134
// store hits in TreeMap - TreeMap does not allow duplicates; existing
135
// entries are silently overwritten
136
for (int hitid = 0; hitid < hitsByRank.length; ++hitid) {
137
resultMap.put(Integer.valueOf(hitsByRank[hitid].doc), // Key: Lucene
139
Integer.valueOf(hitid)); // Value: Hits-Objekt Index
142
// now make a query using the sort criteria
143
ScoreDoc[] resultSort = searcher.search(query, null, Integer.MAX_VALUE,
145
checkHits(resultSort, "Sort by custom criteria: "); // check for duplicates
147
// besides the sorting both sets of hits must be identical
148
for (int hitid = 0; hitid < resultSort.length; ++hitid) {
149
Integer idHitDate = Integer.valueOf(resultSort[hitid].doc); // document ID
152
if (!resultMap.containsKey(idHitDate)) {
153
log("ID " + idHitDate + " not found. Possibliy a duplicate.");
155
assertTrue(resultMap.containsKey(idHitDate)); // same ID must be in the
156
// Map from the rank-sorted
158
// every hit must appear once in both result sets --> remove it from the
160
// At the end the Map must be empty!
161
resultMap.remove(idHitDate);
163
if (resultMap.size() == 0) {
164
// log("All hits matched");
166
log("Couldn't match " + resultMap.size() + " hits.");
168
assertEquals(resultMap.size(), 0);
172
* Check the hits for duplicates.
176
private void checkHits(ScoreDoc[] hits, String prefix) {
178
Map<Integer,Integer> idMap = new TreeMap<Integer,Integer>();
179
for (int docnum = 0; docnum < hits.length; ++docnum) {
180
Integer luceneId = null;
182
luceneId = Integer.valueOf(hits[docnum].doc);
183
if (idMap.containsKey(luceneId)) {
184
StringBuilder message = new StringBuilder(prefix);
185
message.append("Duplicate key for hit index = ");
186
message.append(docnum);
187
message.append(", previous index = ");
188
message.append((idMap.get(luceneId)).toString());
189
message.append(", Lucene ID = ");
190
message.append(luceneId);
191
log(message.toString());
193
idMap.put(luceneId, Integer.valueOf(docnum));
199
// Simply write to console - choosen to be independant of log4j etc
200
private void log(String message) {
201
if (VERBOSE) System.out.println(message);
204
public class CustomSearcher extends IndexSearcher {
205
private int switcher;
210
public CustomSearcher(IndexReader r, int switcher) {
212
this.switcher = switcher;
219
* org.apache.lucene.search.Searchable#search(org.apache.lucene.search.Query
220
* , org.apache.lucene.search.Filter, int, org.apache.lucene.search.Sort)
223
public TopFieldDocs search(Query query, Filter filter, int nDocs, Sort sort)
225
BooleanQuery bq = new BooleanQuery();
226
bq.add(query, BooleanClause.Occur.MUST);
227
bq.add(new TermQuery(new Term("mandant", Integer.toString(switcher))),
228
BooleanClause.Occur.MUST);
229
return super.search(bq, filter, nDocs, sort);
236
* org.apache.lucene.search.Searchable#search(org.apache.lucene.search.Query
237
* , org.apache.lucene.search.Filter, int)
240
public TopDocs search(Query query, Filter filter, int nDocs)
242
BooleanQuery bq = new BooleanQuery();
243
bq.add(query, BooleanClause.Occur.MUST);
244
bq.add(new TermQuery(new Term("mandant", Integer.toString(switcher))),
245
BooleanClause.Occur.MUST);
246
return super.search(bq, filter, nDocs);
250
private class RandomGen {
251
RandomGen(Random random) {
252
this.random = random;
255
private Random random;
256
private Calendar base = new GregorianCalendar(1980, 1, 1);
258
// Just to generate some different Lucene Date strings
259
private String getLuceneDate() {
260
return DateTools.timeToString(base.getTimeInMillis() + random.nextInt()
261
- Integer.MIN_VALUE, DateTools.Resolution.DAY);