1 | /**********************************************************************
|
---|
2 | *
|
---|
3 | * GS2LuceneQuery.java
|
---|
4 | *
|
---|
5 | * Copyright 2004 The New Zealand Digital Library Project
|
---|
6 | *
|
---|
7 | * A component of the Greenstone digital library software
|
---|
8 | * from the New Zealand Digital Library Project at the
|
---|
9 | * University of Waikato, New Zealand.
|
---|
10 | *
|
---|
11 | * This program is free software; you can redistribute it and/or modify
|
---|
12 | * it under the terms of the GNU General Public License as published by
|
---|
13 | * the Free Software Foundation; either version 2 of the License, or
|
---|
14 | * (at your option) any later version.
|
---|
15 | *
|
---|
16 | * This program is distributed in the hope that it will be useful,
|
---|
17 | * but WITHOUT ANY WARRANTY; without even the implied warranty of
|
---|
18 | * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
---|
19 | * GNU General Public License for more details.
|
---|
20 | *
|
---|
21 | * You should have received a copy of the GNU General Public License
|
---|
22 | * along with this program; if not, write to the Free Software
|
---|
23 | * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
|
---|
24 | *
|
---|
25 | *********************************************************************/
|
---|
26 | package org.greenstone.LuceneWrapper4;
|
---|
27 |
|
---|
28 |
|
---|
29 | import java.io.*;
|
---|
30 | import java.util.*;
|
---|
31 | import java.util.regex.*;
|
---|
32 |
|
---|
33 | import org.apache.lucene.analysis.Analyzer;
|
---|
34 | import org.apache.lucene.analysis.standard.StandardAnalyzer;
|
---|
35 | import org.apache.lucene.document.Document;
|
---|
36 | import org.apache.lucene.index.DirectoryReader;
|
---|
37 | import org.apache.lucene.index.IndexReader;
|
---|
38 | import org.apache.lucene.index.Term;
|
---|
39 | //import org.apache.lucene.index.TermDocs;
|
---|
40 | import org.apache.lucene.queryparser.classic.ParseException;
|
---|
41 | import org.apache.lucene.queryparser.classic.QueryParser;
|
---|
42 | import org.apache.lucene.search.BooleanQuery; // for the TooManyClauses exception
|
---|
43 | import org.apache.lucene.search.Filter;
|
---|
44 | import org.apache.lucene.search.IndexSearcher;
|
---|
45 | import org.apache.lucene.search.MultiTermQuery;
|
---|
46 | import org.apache.lucene.search.MultiTermQuery.ConstantScoreAutoRewrite;
|
---|
47 | import org.apache.lucene.search.Query;
|
---|
48 | import org.apache.lucene.search.TermRangeFilter;
|
---|
49 | import org.apache.lucene.search.IndexSearcher; // Searcher is deprecated
|
---|
50 | import org.apache.lucene.search.ScoreDoc;
|
---|
51 | import org.apache.lucene.search.Sort;
|
---|
52 | import org.apache.lucene.search.SortField;
|
---|
53 | import org.apache.lucene.search.TopFieldDocs;
|
---|
54 |
|
---|
55 | import org.apache.lucene.index.DocsEnum;
|
---|
56 | import org.apache.lucene.index.MultiFields;
|
---|
57 |
|
---|
58 | import org.apache.lucene.store.Directory;
|
---|
59 | import org.apache.lucene.store.FSDirectory;
|
---|
60 |
|
---|
61 | import org.apache.lucene.util.Bits;
|
---|
62 | import org.apache.lucene.util.BytesRef;
|
---|
63 | import org.apache.lucene.util.Version;
|
---|
64 |
|
---|
65 | public class GS2LuceneQuery extends SharedSoleneQuery
|
---|
66 | {
|
---|
67 | public static String SORT_RANK = "rank";
|
---|
68 | public static String SORT_NATURAL = "natural";
|
---|
69 |
|
---|
70 | protected String full_indexdir="";
|
---|
71 |
|
---|
72 | protected SortField.Type sort_type = SortField.Type.SCORE;
|
---|
73 | protected boolean reverse_sort = false;
|
---|
74 | protected Sort sorter=new Sort();
|
---|
75 | protected Filter filter = null;
|
---|
76 |
|
---|
77 | protected QueryParser query_parser = null;
|
---|
78 | protected QueryParser query_parser_no_stop_words = null;
|
---|
79 | protected IndexSearcher searcher = null;
|
---|
80 | protected IndexReader reader = null;
|
---|
81 |
|
---|
82 | public GS2LuceneQuery() {
|
---|
83 | super();
|
---|
84 |
|
---|
85 | // Create one query parser with the standard set of stop words, and one with none
|
---|
86 |
|
---|
87 | query_parser = new QueryParser(GSLuceneConstants.MATCH_VERSION, TEXTFIELD, new GS2Analyzer()); // uses built-in stop_words_set
|
---|
88 | query_parser_no_stop_words = new QueryParser(GSLuceneConstants.MATCH_VERSION, TEXTFIELD, new GS2Analyzer(new String[] { }));
|
---|
89 | }
|
---|
90 |
|
---|
91 |
|
---|
92 | public boolean initialise() {
|
---|
93 |
|
---|
94 | if (!super.initialise()) {
|
---|
95 | return false;
|
---|
96 | }
|
---|
97 |
|
---|
98 |
|
---|
99 | if (full_indexdir==null || full_indexdir.length()==-1){
|
---|
100 | utf8out.println("Index directory is not indicated ");
|
---|
101 | utf8out.flush();
|
---|
102 | return false;
|
---|
103 | }
|
---|
104 |
|
---|
105 | try {
|
---|
106 | Directory full_indexdir_dir = FSDirectory.open(new File(full_indexdir));
|
---|
107 |
|
---|
108 | reader = DirectoryReader.open(full_indexdir_dir); // Returns a IndexReader reading the index in the given Directory. now readOnly=true by default, and therefore also for searcher
|
---|
109 | searcher = new IndexSearcher(reader); // during searcher.search() will get it to compute ranks when sorting by fields
|
---|
110 |
|
---|
111 | this.sorter = new Sort(new SortField(this.sort_field, this.sort_type, this.reverse_sort));
|
---|
112 | }
|
---|
113 | catch (IOException exception) {
|
---|
114 | exception.printStackTrace();
|
---|
115 | return false;
|
---|
116 | }
|
---|
117 | return true;
|
---|
118 |
|
---|
119 | }
|
---|
120 |
|
---|
121 | public void setIndexDir(String full_indexdir) {
|
---|
122 | this.full_indexdir = full_indexdir;
|
---|
123 | }
|
---|
124 |
|
---|
125 | public void setSortField(String sort_field) {
|
---|
126 | if (sort_field.equals(SORT_RANK)) {
|
---|
127 | this.sort_field = null;
|
---|
128 | this.sort_type = SortField.Type.SCORE;
|
---|
129 | } else if (sort_field.equals(SORT_NATURAL)) {
|
---|
130 | this.sort_field = null;
|
---|
131 | this.sort_type = SortField.Type.DOC;
|
---|
132 | } else {
|
---|
133 | this.sort_field = sort_field;
|
---|
134 | this.sort_type = SortField.Type.STRING; // for now. numeric??
|
---|
135 | }
|
---|
136 | }
|
---|
137 | public void setReverseSort(boolean reverse) {
|
---|
138 | this.reverse_sort = reverse;
|
---|
139 | }
|
---|
140 | public boolean getReverseSort() {
|
---|
141 | return this.reverse_sort;
|
---|
142 | }
|
---|
143 |
|
---|
144 | public void setFilterString(String filter_string) {
|
---|
145 | super.setFilterString(filter_string);
|
---|
146 | this.filter = parseFilterString(filter_string);
|
---|
147 | }
|
---|
148 |
|
---|
149 | public Filter getFilter() {
|
---|
150 | return this.filter;
|
---|
151 | }
|
---|
152 |
|
---|
153 |
|
---|
154 | public LuceneQueryResult runQuery(String query_string) {
|
---|
155 |
|
---|
156 | if (query_string == null || query_string.equals("")) {
|
---|
157 | utf8out.println("The query word is not indicated ");
|
---|
158 | utf8out.flush();
|
---|
159 | return null;
|
---|
160 | }
|
---|
161 |
|
---|
162 | LuceneQueryResult lucene_query_result=new LuceneQueryResult();
|
---|
163 | lucene_query_result.clear();
|
---|
164 |
|
---|
165 | try {
|
---|
166 | Query query_including_stop_words = query_parser_no_stop_words.parse(query_string);
|
---|
167 | query_including_stop_words = query_including_stop_words.rewrite(reader);
|
---|
168 |
|
---|
169 | // System.err.println("********* query_string " + query_string + "****");
|
---|
170 |
|
---|
171 | Query query = parseQuery(reader, query_parser, query_string, fuzziness);
|
---|
172 |
|
---|
173 | // GS2's LuceneWrapper uses lucene-2.3.2. GS3's LuceneWrapper3 works with lucene-3.3.0.
|
---|
174 | // This change in lucene core library for GS3 (present since after version 2.4.1) had the
|
---|
175 | // side-effect that searching on "econom*" didn't display what terms it was searching for,
|
---|
176 | // whereas it had done so in GS2.
|
---|
177 |
|
---|
178 | // The details of this problem and its current solution are explained in the ticket
|
---|
179 | // http://trac.greenstone.org/ticket/845
|
---|
180 |
|
---|
181 | // We need to change the settings for the rewriteMethod in order to get searches on wildcards
|
---|
182 | // to produce search terms again when the query gets rewritten.
|
---|
183 |
|
---|
184 | // We try, in order:
|
---|
185 | // 1. RewriteMethod set to BooleanQuery, to get it working as in GS2 which uses lucene-2.3.2
|
---|
186 | // it will expand wildcard searches to its terms when searching at both section AND doc level.
|
---|
187 | // If that throws a TooManyClauses exception (like when searching for "a*" over lucene demo collection)
|
---|
188 | // 2. Then try a custom rewriteMethod which sets termCountCutoff=350 and docCountPercent cutoff=0.1%
|
---|
189 | // If that throws a TooManyClauses exception (could perhaps happen if the collection has a huge number of docs
|
---|
190 | // 3. Then try the default apache rewriteMethod with its optimum defaults of
|
---|
191 | // termCountCutoff=350 and docCountPercent cutoff=0.1%
|
---|
192 | // See http://lucene.apache.org/core/3_6_1/api/core/org/apache/lucene/search/MultiTermQuery.html
|
---|
193 |
|
---|
194 | if(query instanceof MultiTermQuery) {
|
---|
195 | MultiTermQuery multiTermQuery = (MultiTermQuery)query;
|
---|
196 | multiTermQuery.setRewriteMethod(MultiTermQuery.CONSTANT_SCORE_BOOLEAN_QUERY_REWRITE);
|
---|
197 | // less CPU intensive than MultiTermQuery.SCORING_BOOLEAN_QUERY_REWRITE)
|
---|
198 | }
|
---|
199 |
|
---|
200 | try {
|
---|
201 | query = query.rewrite(reader);
|
---|
202 | }
|
---|
203 | catch(BooleanQuery.TooManyClauses clauseException) {
|
---|
204 | // Example test case: try searching the lucene demo collection for "a*"
|
---|
205 | // and you'll hit this exception
|
---|
206 |
|
---|
207 | lucene_query_result.setError(LuceneQueryResult.TOO_MANY_CLAUSES_ERROR);
|
---|
208 |
|
---|
209 | if(query instanceof MultiTermQuery) {
|
---|
210 |
|
---|
211 | // CustomRewriteMethod: setting the docCountPercent cutoff to a custom 100%.
|
---|
212 | // This will at least expand the query to its terms when searching with wildcards at section-level
|
---|
213 | // (though it doesn't seem to work for doc-level searches, no matter what the cutoffs are set to).
|
---|
214 |
|
---|
215 | MultiTermQuery.ConstantScoreAutoRewrite customRewriteMethod = new MultiTermQuery.ConstantScoreAutoRewrite();
|
---|
216 | customRewriteMethod.setDocCountPercent(100.0);
|
---|
217 | customRewriteMethod.setTermCountCutoff(350); // same as default
|
---|
218 |
|
---|
219 | MultiTermQuery multiTermQuery = (MultiTermQuery)query;
|
---|
220 | multiTermQuery.setRewriteMethod(customRewriteMethod);
|
---|
221 | try {
|
---|
222 | query = query.rewrite(reader);
|
---|
223 | }
|
---|
224 | catch(BooleanQuery.TooManyClauses clauseExceptionAgain) {
|
---|
225 |
|
---|
226 | // do what the code originally did: use the default rewriteMethod which
|
---|
227 | // uses a default docCountPercent=0.1 (%) and termCountCutoff=350
|
---|
228 |
|
---|
229 | multiTermQuery.setRewriteMethod(MultiTermQuery.CONSTANT_SCORE_AUTO_REWRITE_DEFAULT);
|
---|
230 | query = query.rewrite(reader);
|
---|
231 | }
|
---|
232 | }
|
---|
233 | }
|
---|
234 |
|
---|
235 | // http://stackoverflow.com/questions/13537126/term-frequency-in-lucene-4-0
|
---|
236 | // http://stackoverflow.com/questions/20575254/lucene-4-4-how-to-get-term-frequency-over-all-index
|
---|
237 | // http://stackoverflow.com/questions/8938960/how-to-get-document-ids-for-document-term-vector-in-lucene?rq=1
|
---|
238 | // https://github.com/hibernate/hibernate-search/blob/master/orm/src/test/java/org/hibernate/search/test/filter/BestDriversFilter.java
|
---|
239 | // http://lucene.apache.org/core/4_7_2/MIGRATE.html
|
---|
240 |
|
---|
241 | // Get the list of expanded query terms and their frequencies
|
---|
242 | // num docs matching, and total frequency
|
---|
243 | HashSet terms = new HashSet();
|
---|
244 | query.extractTerms(terms);
|
---|
245 |
|
---|
246 | HashMap doc_term_freq_map = new HashMap();
|
---|
247 |
|
---|
248 | Iterator iter = terms.iterator();
|
---|
249 |
|
---|
250 | Bits liveDocs = null;
|
---|
251 | if(reader.hasDeletions()) {
|
---|
252 | System.err.println("@@@ GS2LuceneQuery.java: There have been deletions. Merging to get liveDocs.");
|
---|
253 | liveDocs = MultiFields.getLiveDocs(reader); // SLOW! But getLiveDocs returns null if there are no deletions
|
---|
254 | }
|
---|
255 |
|
---|
256 | while (iter.hasNext()) {
|
---|
257 |
|
---|
258 | // http://stackoverflow.com/questions/13537126/term-frequency-in-lucene-4-0
|
---|
259 |
|
---|
260 | Term term = (Term) iter.next();
|
---|
261 | BytesRef term_bytes = term.bytes();
|
---|
262 | DocsEnum term_docs = MultiFields.getTermDocsEnum(reader, liveDocs, term.field(), term_bytes); // flags?
|
---|
263 |
|
---|
264 | // Get the term frequency over all the documents
|
---|
265 | //TermDocs term_docs = reader.termDocs(term);
|
---|
266 | int term_freq = 0;
|
---|
267 | int match_docs = 0;
|
---|
268 |
|
---|
269 | if(term_docs != null) {
|
---|
270 | int docID = -1;
|
---|
271 | while((docID = term_docs.nextDoc()) != DocsEnum.NO_MORE_DOCS) {//while (term_docs.next())
|
---|
272 | if (term_docs.freq() != 0)
|
---|
273 | {
|
---|
274 | term_freq += term_docs.freq();
|
---|
275 | match_docs++;
|
---|
276 |
|
---|
277 | // Calculate the document-level term frequency as well
|
---|
278 | Integer lucene_doc_num_obj = new Integer(term_docs.docID());
|
---|
279 | int doc_term_freq = 0;
|
---|
280 | if (doc_term_freq_map.containsKey(lucene_doc_num_obj))
|
---|
281 | {
|
---|
282 | doc_term_freq = ((Integer) doc_term_freq_map.get(lucene_doc_num_obj)).intValue();
|
---|
283 | }
|
---|
284 | doc_term_freq += term_docs.freq();
|
---|
285 |
|
---|
286 | doc_term_freq_map.put(lucene_doc_num_obj, new Integer(doc_term_freq));
|
---|
287 | }
|
---|
288 | }
|
---|
289 | } else {
|
---|
290 | System.err.println("@@@ GS2LuceneQuery.java: term_docs is null for term " + term.text());
|
---|
291 | }
|
---|
292 |
|
---|
293 | // Create a term
|
---|
294 | lucene_query_result.addTerm(term.text(), term.field(), match_docs, term_freq);
|
---|
295 | }
|
---|
296 |
|
---|
297 | // Get the list of stop words removed from the query
|
---|
298 | HashSet terms_including_stop_words = new HashSet();
|
---|
299 | query_including_stop_words.extractTerms(terms_including_stop_words);
|
---|
300 | Iterator terms_including_stop_words_iter = terms_including_stop_words.iterator();
|
---|
301 | while (terms_including_stop_words_iter.hasNext()) {
|
---|
302 | Term term = (Term) terms_including_stop_words_iter.next();
|
---|
303 | if (!terms.contains(term)) {
|
---|
304 | lucene_query_result.addStopWord(term.text());
|
---|
305 | }
|
---|
306 | }
|
---|
307 |
|
---|
308 | // Extracting all documents for a given search - http://www.gossamer-threads.com/lists/lucene/java-user/134873
|
---|
309 | // http://lucene.apache.org/core/3_4_0/api/core/org/apache/lucene/search/TotalHitCountCollector.html
|
---|
310 | // http://lucene.apache.org/core/4_7_2/core/index.html?org/apache/lucene/search/TopFieldDocs.html
|
---|
311 |
|
---|
312 | // 1. Figure out how many results there will be.
|
---|
313 | //TotalHitCountCollecter countCollector = new TotalHitCountCollector();
|
---|
314 | //searcher.search(query, filter, collector);
|
---|
315 | //int hitCount = collector.count;
|
---|
316 |
|
---|
317 | // Actually do the query
|
---|
318 | // Simple case for getting all the matching documents
|
---|
319 | if (end_results == Integer.MAX_VALUE) {
|
---|
320 | // Perform the query (filter and sorter may be null)
|
---|
321 | TopFieldDocs hits = searcher.search(query, filter, end_results, sorter, true, true); // doDocScores=true, doMaxScore=true
|
---|
322 | // Is there a slight difference in the definition between
|
---|
323 | // https://lucene.apache.org/core/3_6_0/api/all/org/apache/lucene/search/IndexSearcher.html#setDefaultFieldSortScoring%28boolean,%20boolean%29
|
---|
324 | // and http://lucene.apache.org/core/4_7_2/core/org/apache/lucene/search/IndexSearcher.html#search%28org.apache.lucene.search.Query,%20org.apache.lucene.search.Filter,%20int,%20org.apache.lucene.search.Sort,%20boolean,%20boolean%29
|
---|
325 | // Seems to be okay.
|
---|
326 | // See also http://stackoverflow.com/questions/7910241/in-lucene-what-is-the-purpose-of-setdefaultfieldsortscoring
|
---|
327 |
|
---|
328 | lucene_query_result.setTotalDocs(hits.totalHits);
|
---|
329 |
|
---|
330 | // Output the matching documents
|
---|
331 | lucene_query_result.setStartResults(start_results);
|
---|
332 | lucene_query_result.setEndResults(hits.totalHits); // ??
|
---|
333 |
|
---|
334 | for (int i = start_results; i < hits.totalHits; i++) {
|
---|
335 | int lucene_doc_num = hits.scoreDocs[i ].doc; // i-1
|
---|
336 | Document doc = reader.document(lucene_doc_num);
|
---|
337 | int doc_term_freq = 0;
|
---|
338 | Integer doc_term_freq_object = (Integer) doc_term_freq_map.get(new Integer(lucene_doc_num));
|
---|
339 | if (doc_term_freq_object != null)
|
---|
340 | {
|
---|
341 | doc_term_freq = doc_term_freq_object.intValue();
|
---|
342 | }
|
---|
343 | lucene_query_result.addDoc(doc.get("docOID").trim(), hits.scoreDocs[i].score, doc_term_freq);
|
---|
344 | }
|
---|
345 | }
|
---|
346 |
|
---|
347 | // Slightly more complicated case for returning a subset of the matching documents
|
---|
348 | else {
|
---|
349 | // Perform the query (filter may be null)
|
---|
350 | TopFieldDocs hits = searcher.search(query, filter, end_results, sorter, true, true); // doDocScores=true, doMaxScore=true
|
---|
351 | // See also http://stackoverflow.com/questions/7910241/in-lucene-what-is-the-purpose-of-setdefaultfieldsortscoring
|
---|
352 | lucene_query_result.setTotalDocs(hits.totalHits);
|
---|
353 |
|
---|
354 | lucene_query_result.setStartResults(start_results);
|
---|
355 | lucene_query_result.setEndResults(end_results < hits.scoreDocs.length ? end_results: hits.scoreDocs.length);
|
---|
356 |
|
---|
357 | // Output the matching documents
|
---|
358 | for (int i = start_results; (i < hits.scoreDocs.length && i < end_results); i++) {
|
---|
359 | int lucene_doc_num = hits.scoreDocs[i].doc;
|
---|
360 | Document doc = reader.document(lucene_doc_num);
|
---|
361 | int doc_term_freq = 0;
|
---|
362 | Integer doc_term_freq_object = (Integer) doc_term_freq_map.get(new Integer(lucene_doc_num));
|
---|
363 | if (doc_term_freq_object != null)
|
---|
364 | {
|
---|
365 | doc_term_freq = doc_term_freq_object.intValue();
|
---|
366 | }
|
---|
367 | lucene_query_result.addDoc(doc.get("docOID").trim(), hits.scoreDocs[i].score, doc_term_freq);
|
---|
368 | }
|
---|
369 | }
|
---|
370 | }
|
---|
371 |
|
---|
372 | catch (ParseException parse_exception) {
|
---|
373 | lucene_query_result.setError(LuceneQueryResult.PARSE_ERROR);
|
---|
374 | }
|
---|
375 | catch (BooleanQuery.TooManyClauses too_many_clauses_exception) {
|
---|
376 | lucene_query_result.setError(LuceneQueryResult.TOO_MANY_CLAUSES_ERROR);
|
---|
377 | }
|
---|
378 | catch (IOException exception) {
|
---|
379 | lucene_query_result.setError(LuceneQueryResult.IO_ERROR);
|
---|
380 | exception.printStackTrace();
|
---|
381 | }
|
---|
382 | catch (Exception exception) {
|
---|
383 | lucene_query_result.setError(LuceneQueryResult.OTHER_ERROR);
|
---|
384 | exception.printStackTrace();
|
---|
385 | }
|
---|
386 | return lucene_query_result;
|
---|
387 | }
|
---|
388 |
|
---|
389 | public void setDefaultConjunctionOperator(String default_conjunction_operator) {
|
---|
390 | super.setDefaultConjunctionOperator(default_conjunction_operator);
|
---|
391 |
|
---|
392 | if (default_conjunction_operator.equals("AND")) {
|
---|
393 | query_parser.setDefaultOperator(query_parser.AND_OPERATOR);
|
---|
394 | query_parser_no_stop_words.setDefaultOperator(query_parser.AND_OPERATOR);
|
---|
395 | } else { // default is OR
|
---|
396 | query_parser.setDefaultOperator(query_parser.OR_OPERATOR);
|
---|
397 | query_parser_no_stop_words.setDefaultOperator(query_parser.OR_OPERATOR);
|
---|
398 | }
|
---|
399 | }
|
---|
400 |
|
---|
401 |
|
---|
402 | public void cleanUp() {
|
---|
403 | super.cleanUp();
|
---|
404 | try {
|
---|
405 | if(reader != null) {
|
---|
406 | reader.close();
|
---|
407 | // Closes files associated with this index. Also saves any new deletions to disk.
|
---|
408 | // No other methods should be called after this has been called.
|
---|
409 | }
|
---|
410 | } catch (IOException exception) {
|
---|
411 | exception.printStackTrace();
|
---|
412 | }
|
---|
413 | }
|
---|
414 |
|
---|
415 |
|
---|
416 | protected Query parseQuery(IndexReader reader, QueryParser query_parser, String query_string, String fuzziness)
|
---|
417 | throws java.io.IOException, org.apache.lucene.queryparser.classic.ParseException
|
---|
418 | {
|
---|
419 | // Split query string into the search terms and the filter terms
|
---|
420 | // * The first +(...) term contains the search terms so count
|
---|
421 | // up '(' and stop when we finish matching ')'
|
---|
422 | int offset = 0;
|
---|
423 | int paren_count = 0;
|
---|
424 | boolean seen_paren = false;
|
---|
425 | while (offset < query_string.length() && (!seen_paren || paren_count > 0)) {
|
---|
426 | if (query_string.charAt(offset) == '(') {
|
---|
427 | paren_count++;
|
---|
428 | seen_paren = true;
|
---|
429 | }
|
---|
430 | if (query_string.charAt(offset) == ')') {
|
---|
431 | paren_count--;
|
---|
432 | }
|
---|
433 | offset++;
|
---|
434 | }
|
---|
435 | String query_prefix = query_string.substring(0, offset);
|
---|
436 | String query_suffix = query_string.substring(offset);
|
---|
437 |
|
---|
438 | ///ystem.err.println("Prefix: " + query_prefix);
|
---|
439 | ///ystem.err.println("Suffix: " + query_suffix);
|
---|
440 |
|
---|
441 | Query query = query_parser.parse(query_prefix);
|
---|
442 | query = query.rewrite(reader);
|
---|
443 |
|
---|
444 | // If this is a fuzzy search, then we need to add the fuzzy
|
---|
445 | // flag to each of the query terms
|
---|
446 | if (fuzziness != null && query.toString().length() > 0) {
|
---|
447 |
|
---|
448 | // Revert the query to a string
|
---|
449 | System.err.println("Rewritten query: " + query.toString());
|
---|
450 | // Search through the string for TX:<term> query terms
|
---|
451 | // and append the ~ operator. Note that this search will
|
---|
452 | // not change phrase searches (TX:"<term> <term>") as
|
---|
453 | // fuzzy searching is not possible for these entries.
|
---|
454 | // Yahoo! Time for a state machine!
|
---|
455 | StringBuffer mutable_query_string = new StringBuffer(query.toString());
|
---|
456 | int o = 0; // Offset
|
---|
457 | // 0 = BASE, 1 = SEEN_T, 2 = SEEN_TX, 3 = SEEN_TX:
|
---|
458 | int s = 0; // State
|
---|
459 | while(o < mutable_query_string.length()) {
|
---|
460 | char c = mutable_query_string.charAt(o);
|
---|
461 | if (s == 0 && c == TEXTFIELD.charAt(0)) {
|
---|
462 | ///ystem.err.println("Found T!");
|
---|
463 | s = 1;
|
---|
464 | }
|
---|
465 | else if (s == 1) {
|
---|
466 | if (c == TEXTFIELD.charAt(1)) {
|
---|
467 | ///ystem.err.println("Found X!");
|
---|
468 | s = 2;
|
---|
469 | }
|
---|
470 | else {
|
---|
471 | s = 0; // Reset
|
---|
472 | }
|
---|
473 | }
|
---|
474 | else if (s == 2) {
|
---|
475 | if (c == ':') {
|
---|
476 | ///ystem.err.println("Found TX:!");
|
---|
477 | s = 3;
|
---|
478 | }
|
---|
479 | else {
|
---|
480 | s = 0; // Reset
|
---|
481 | }
|
---|
482 | }
|
---|
483 | else if (s == 3) {
|
---|
484 | // Don't process phrases
|
---|
485 | if (c == '"') {
|
---|
486 | ///ystem.err.println("Stupid phrase...");
|
---|
487 | s = 0; // Reset
|
---|
488 | }
|
---|
489 | // Found the end of the term... add the
|
---|
490 | // fuzzy search indicator
|
---|
491 | // Nor outside the scope of parentheses
|
---|
492 | else if (Character.isWhitespace(c) || c == ')') {
|
---|
493 | ///ystem.err.println("Yahoo! Found fuzzy term.");
|
---|
494 | mutable_query_string.insert(o, '~' + fuzziness);
|
---|
495 | o++;
|
---|
496 | s = 0; // Reset
|
---|
497 | }
|
---|
498 | }
|
---|
499 | o++;
|
---|
500 | }
|
---|
501 | // If we were in the state of looking for the end of a
|
---|
502 | // term - then we just found it!
|
---|
503 | if (s == 3) {
|
---|
504 |
|
---|
505 | mutable_query_string.append('~' + fuzziness);
|
---|
506 | }
|
---|
507 | // Reparse the query
|
---|
508 | ///ystem.err.println("Fuzzy query: " + mutable_query_string.toString() + query_suffix);
|
---|
509 | query = query_parser.parse(mutable_query_string.toString() + query_suffix);
|
---|
510 | }
|
---|
511 | else {
|
---|
512 | query = query_parser.parse(query_prefix + query_suffix);
|
---|
513 | }
|
---|
514 |
|
---|
515 | return query;
|
---|
516 | }
|
---|
517 |
|
---|
518 | protected Filter parseFilterString(String filter_string)
|
---|
519 | {
|
---|
520 | Filter result = null;
|
---|
521 | Pattern pattern = Pattern.compile("\\s*\\+(\\w+)\\:([\\{\\[])(\\d+)\\s+TO\\s+(\\d+)([\\}\\]])\\s*");
|
---|
522 | Matcher matcher = pattern.matcher(filter_string);
|
---|
523 | if (matcher.matches()) {
|
---|
524 | String field_name = matcher.group(1);
|
---|
525 | boolean include_lower = matcher.group(2).equals("[");
|
---|
526 | BytesRef lower_term = new BytesRef(matcher.group(3));
|
---|
527 | BytesRef upper_term = new BytesRef(matcher.group(4));
|
---|
528 | boolean include_upper = matcher.group(5).equals("]");
|
---|
529 | result = new TermRangeFilter(field_name, lower_term, upper_term, include_lower, include_upper);
|
---|
530 | }
|
---|
531 | else {
|
---|
532 | System.err.println("Error: Could not understand filter string \"" + filter_string + "\"");
|
---|
533 | }
|
---|
534 | return result;
|
---|
535 | }
|
---|
536 |
|
---|
537 |
|
---|
538 | /** command line program and auxiliary methods */
|
---|
539 |
|
---|
540 | // Fairly self-explanatory I should hope
|
---|
541 | static protected boolean query_result_caching_enabled = false;
|
---|
542 |
|
---|
543 |
|
---|
544 | static public void main (String args[])
|
---|
545 | {
|
---|
546 | if (args.length == 0) {
|
---|
547 | System.out.println("Usage: GS2LuceneQuery <index directory> [-fuzziness value] [-filter filter_string] [-sort sort_field] [-reverse_sort][-dco AND|OR] [-startresults number -endresults number] [query]");
|
---|
548 | return;
|
---|
549 | }
|
---|
550 |
|
---|
551 | try {
|
---|
552 | String index_directory = args[0];
|
---|
553 |
|
---|
554 | GS2LuceneQuery queryer = new GS2LuceneQuery();
|
---|
555 | queryer.setIndexDir(index_directory);
|
---|
556 |
|
---|
557 | // Prepare the index cache directory, if query result caching is enabled
|
---|
558 | if (query_result_caching_enabled) {
|
---|
559 | // Make the index cache directory if it doesn't already exist
|
---|
560 | File index_cache_directory = new File(index_directory, "cache");
|
---|
561 | if (!index_cache_directory.exists()) {
|
---|
562 | index_cache_directory.mkdir();
|
---|
563 | }
|
---|
564 |
|
---|
565 | // Disable caching if the index cache directory isn't available
|
---|
566 | if (!index_cache_directory.exists() || !index_cache_directory.isDirectory()) {
|
---|
567 | query_result_caching_enabled = false;
|
---|
568 | }
|
---|
569 | }
|
---|
570 |
|
---|
571 | String query_string = null;
|
---|
572 |
|
---|
573 | // Parse the command-line arguments
|
---|
574 | for (int i = 1; i < args.length; i++) {
|
---|
575 | if (args[i].equals("-sort")) {
|
---|
576 | i++;
|
---|
577 | queryer.setSortField(args[i]);
|
---|
578 | }
|
---|
579 | else if (args[i].equals("-reverse_sort")) {
|
---|
580 | queryer.setReverseSort(true);
|
---|
581 | }
|
---|
582 | else if (args[i].equals("-filter")) {
|
---|
583 | i++;
|
---|
584 | queryer.setFilterString(args[i]);
|
---|
585 | }
|
---|
586 | else if (args[i].equals("-dco")) {
|
---|
587 | i++;
|
---|
588 | queryer.setDefaultConjunctionOperator(args[i]);
|
---|
589 | }
|
---|
590 | else if (args[i].equals("-fuzziness")) {
|
---|
591 | i++;
|
---|
592 | queryer.setFuzziness(args[i]);
|
---|
593 | }
|
---|
594 | else if (args[i].equals("-startresults")) {
|
---|
595 | i++;
|
---|
596 | if (args[i].matches("\\d+")) {
|
---|
597 | queryer.setStartResults(Integer.parseInt(args[i]));
|
---|
598 | }
|
---|
599 | }
|
---|
600 | else if (args[i].equals("-endresults")) {
|
---|
601 | i++;
|
---|
602 | if (args[i].matches("\\d+")) {
|
---|
603 | queryer.setEndResults(Integer.parseInt(args[i]));
|
---|
604 | }
|
---|
605 | }
|
---|
606 | else {
|
---|
607 | query_string = args[i];
|
---|
608 | }
|
---|
609 | }
|
---|
610 |
|
---|
611 | if (!queryer.initialise()) {
|
---|
612 | return;
|
---|
613 | }
|
---|
614 |
|
---|
615 | // The query string has been specified as a command-line argument
|
---|
616 | if (query_string != null) {
|
---|
617 | runQueryCaching(index_directory, queryer, query_string);
|
---|
618 | }
|
---|
619 |
|
---|
620 | // Read queries from STDIN
|
---|
621 | else {
|
---|
622 | BufferedReader in = new BufferedReader(new InputStreamReader(System.in, "UTF-8"));
|
---|
623 | while (true) {
|
---|
624 | // Read the query from STDIN
|
---|
625 | query_string = in.readLine();
|
---|
626 | if (query_string == null || query_string.length() == -1) {
|
---|
627 | break;
|
---|
628 | }
|
---|
629 |
|
---|
630 | runQueryCaching(index_directory, queryer, query_string);
|
---|
631 |
|
---|
632 | }
|
---|
633 | }
|
---|
634 | queryer.cleanUp();
|
---|
635 | }
|
---|
636 | catch (IOException exception) {
|
---|
637 | exception.printStackTrace();
|
---|
638 | }
|
---|
639 | }
|
---|
640 |
|
---|
641 | protected static void runQueryCaching(String index_directory, GS2LuceneQuery queryer, String query_string)
|
---|
642 | throws IOException
|
---|
643 | {
|
---|
644 | StringBuffer query_results_xml = new StringBuffer();
|
---|
645 |
|
---|
646 | // Check if this query result has been cached from a previous search (if it's enabled)
|
---|
647 | File query_result_cache_file = null;
|
---|
648 | if (query_result_caching_enabled) {
|
---|
649 | // Generate the cache file name from the query options
|
---|
650 | String query_result_cache_file_name = query_string + "-";
|
---|
651 | String fuzziness = queryer.getFuzziness();
|
---|
652 | query_result_cache_file_name += ((fuzziness != null) ? fuzziness : "") + "-";
|
---|
653 | String filter_string = queryer.getFilterString();
|
---|
654 | query_result_cache_file_name += ((filter_string != null) ? filter_string : "") + "-";
|
---|
655 | String sort_string = queryer.getSortField();
|
---|
656 | query_result_cache_file_name += ((sort_string != null) ? sort_string : "") + "-";
|
---|
657 | String reverse_sort_string = (queryer.getReverseSort() ? "1" : "0");
|
---|
658 | query_result_cache_file_name += reverse_sort_string + "-";
|
---|
659 | String default_conjunction_operator = queryer.getDefaultConjunctionOperator();
|
---|
660 | query_result_cache_file_name += default_conjunction_operator + "-";
|
---|
661 | int start_results = queryer.getStartResults();
|
---|
662 | int end_results = queryer.getEndResults();
|
---|
663 | query_result_cache_file_name += start_results + "-" + end_results;
|
---|
664 | query_result_cache_file_name = fileSafe(query_result_cache_file_name);
|
---|
665 |
|
---|
666 | // If the query result cache file exists, just return its contents and we're done
|
---|
667 | File index_cache_directory = new File(index_directory, "cache");
|
---|
668 | query_result_cache_file = new File(index_cache_directory, query_result_cache_file_name);
|
---|
669 | if (query_result_cache_file.exists() && query_result_cache_file.isFile()) {
|
---|
670 | FileInputStream fis = new FileInputStream(query_result_cache_file);
|
---|
671 | InputStreamReader isr = new InputStreamReader(fis, "UTF-8");
|
---|
672 | BufferedReader buffered_reader = new BufferedReader(isr);
|
---|
673 | String line = "";
|
---|
674 | while ((line = buffered_reader.readLine()) != null) {
|
---|
675 | query_results_xml.append(line + "\n");
|
---|
676 | }
|
---|
677 | String query_results_xml_string = query_results_xml.toString();
|
---|
678 | query_results_xml_string = query_results_xml_string.replaceFirst("cached=\"false\"", "cached=\"true\"");
|
---|
679 |
|
---|
680 | utf8out.print(query_results_xml_string);
|
---|
681 | utf8out.flush();
|
---|
682 |
|
---|
683 | return;
|
---|
684 | }
|
---|
685 | }
|
---|
686 |
|
---|
687 | // not cached
|
---|
688 | query_results_xml.append("<ResultSet cached=\"false\">\n");
|
---|
689 | query_results_xml.append("<QueryString>" + LuceneQueryResult.xmlSafe(query_string) + "</QueryString>\n");
|
---|
690 | Filter filter = queryer.getFilter();
|
---|
691 | if (filter != null) {
|
---|
692 | query_results_xml.append("<FilterString>" + filter.toString() + "</FilterString>\n");
|
---|
693 | }
|
---|
694 |
|
---|
695 | LuceneQueryResult query_result = queryer.runQuery(query_string);
|
---|
696 | if (query_result == null) {
|
---|
697 | System.err.println("Couldn't run the query");
|
---|
698 | return;
|
---|
699 | }
|
---|
700 |
|
---|
701 | if (query_result.getError() != LuceneQueryResult.NO_ERROR) {
|
---|
702 | query_results_xml.append("<Error type=\""+query_result.getErrorString()+"\" />\n");
|
---|
703 | } else {
|
---|
704 | query_results_xml.append(query_result.getXMLString());
|
---|
705 | }
|
---|
706 | query_results_xml.append("</ResultSet>\n");
|
---|
707 |
|
---|
708 | utf8out.print(query_results_xml);
|
---|
709 | utf8out.flush();
|
---|
710 |
|
---|
711 | // Cache this query result, if desired
|
---|
712 | if (query_result_caching_enabled) {
|
---|
713 | // Catch any exceptions thrown trying to write the query result cache file and warn about them, but don't
|
---|
714 | // bother with the full stack trace. It won't affect the functionality if we can't write some cache
|
---|
715 | // files, it will just affect the speed of subsequent requests.
|
---|
716 | // Example exceptions are "permission denied" errors, or "filename too long" errors (the filter string
|
---|
717 | // can get very long in some collections)
|
---|
718 | try
|
---|
719 | {
|
---|
720 | FileWriter query_result_cache_file_writer = new FileWriter(query_result_cache_file);
|
---|
721 | query_result_cache_file_writer.write(query_results_xml.toString());
|
---|
722 | query_result_cache_file_writer.close();
|
---|
723 | }
|
---|
724 | catch (Exception exception)
|
---|
725 | {
|
---|
726 | System.err.println("Warning: Exception occurred trying to write query result cache file (" + exception + ")");
|
---|
727 | }
|
---|
728 | }
|
---|
729 | }
|
---|
730 |
|
---|
731 | protected static String fileSafe(String text)
|
---|
732 | {
|
---|
733 | StringBuffer file_safe_text = new StringBuffer();
|
---|
734 | for (int i = 0; i < text.length(); i++) {
|
---|
735 | char character = text.charAt(i);
|
---|
736 | if ((character >= 'A' && character <= 'Z') || (character >= 'a' && character <= 'z') || (character >= '0' && character <= '9') || character == '-') {
|
---|
737 | file_safe_text.append(character);
|
---|
738 | }
|
---|
739 | else {
|
---|
740 | file_safe_text.append('%');
|
---|
741 | file_safe_text.append((int) character);
|
---|
742 | }
|
---|
743 | }
|
---|
744 | return file_safe_text.toString();
|
---|
745 | }
|
---|
746 |
|
---|
747 |
|
---|
748 | }
|
---|
749 |
|
---|
750 |
|
---|