1 | /**********************************************************************
|
---|
2 | *
|
---|
3 | * lucenesearch.cpp --
|
---|
4 | * Copyright (C) 1999-2002 The New Zealand Digital Library Project
|
---|
5 | *
|
---|
6 | * A component of the Greenstone digital library software
|
---|
7 | * from the New Zealand Digital Library Project at the
|
---|
8 | * University of Waikato, New Zealand.
|
---|
9 | *
|
---|
10 | * This program is free software; you can redistribute it and/or modify
|
---|
11 | * it under the terms of the GNU General Public License as published by
|
---|
12 | * the Free Software Foundation; either version 2 of the License, or
|
---|
13 | * (at your option) any later version.
|
---|
14 | *
|
---|
15 | * This program is distributed in the hope that it will be useful,
|
---|
16 | * but WITHOUT ANY WARRANTY; without even the implied warranty of
|
---|
17 | * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
---|
18 | * GNU General Public License for more details.
|
---|
19 | *
|
---|
20 | * You should have received a copy of the GNU General Public License
|
---|
21 | * along with this program; if not, write to the Free Software
|
---|
22 | * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
|
---|
23 | *
|
---|
24 | *********************************************************************/
|
---|
25 |
|
---|
26 |
|
---|
27 | #if defined(GSDL_USE_OBJECTSPACE)
|
---|
28 | # include <ospace\std\iostream>
|
---|
29 | #elif defined(GSDL_USE_IOS_H)
|
---|
30 | # include <iostream.h>
|
---|
31 | #else
|
---|
32 | # include <iostream>
|
---|
33 | #endif
|
---|
34 |
|
---|
35 | #include <stdio.h>
|
---|
36 |
|
---|
37 | #include "gsdlconf.h"
|
---|
38 | #include "lucenesearch.h"
|
---|
39 | #include "fileutil.h"
|
---|
40 | #include "queryinfo.h"
|
---|
41 | #include "gsdlunicode.h"
|
---|
42 |
|
---|
43 | #include "expat_resultset.h"
|
---|
44 |
|
---|
45 |
|
---|
46 | static text_t getindexsuffix(const queryparamclass &qp) {
|
---|
47 | text_t indexsuffix = "index";
|
---|
48 | // get the first char of the level to be the start of the index name
|
---|
49 | text_t suffix = substr(qp.level.begin(), qp.level.begin()+1);
|
---|
50 | lc(suffix);
|
---|
51 | text_t ind = qp.index;
|
---|
52 | text_t sub = qp.subcollection;
|
---|
53 | text_t lang = qp.language;
|
---|
54 |
|
---|
55 | // collection name not added for Lucene
|
---|
56 | indexsuffix = filename_cat(indexsuffix, suffix +ind + sub + lang);
|
---|
57 | return indexsuffix;
|
---|
58 |
|
---|
59 | }
|
---|
60 |
|
---|
61 | ////////////////////
|
---|
62 | // lucenesearch class //
|
---|
63 | ////////////////////
|
---|
64 |
|
---|
65 | lucenesearchclass::lucenesearchclass ()
|
---|
66 | : searchclass() {
|
---|
67 |
|
---|
68 | gdbm_level = "Doc";
|
---|
69 | }
|
---|
70 |
|
---|
71 | lucenesearchclass::~lucenesearchclass ()
|
---|
72 | {
|
---|
73 | if (cache != NULL)
|
---|
74 | {
|
---|
75 | delete cache;
|
---|
76 | cache = NULL;
|
---|
77 | }
|
---|
78 | }
|
---|
79 |
|
---|
80 | void lucenesearchclass::set_gdbm_level(const text_t &level) {
|
---|
81 | gdbm_level = level;
|
---|
82 |
|
---|
83 | }
|
---|
84 |
|
---|
85 |
|
---|
86 | bool lucenesearchclass::search(const queryparamclass &queryparams,
|
---|
87 | queryresultsclass &queryresult) {
|
---|
88 |
|
---|
89 | #ifdef __WIN32__
|
---|
90 | char basepath[]="";
|
---|
91 | #else
|
---|
92 | char basepath[] = "/";
|
---|
93 | #endif
|
---|
94 |
|
---|
95 | cerr << "**** in luecen search" << endl;
|
---|
96 |
|
---|
97 | char *indexname = (filename_cat(collectdir, getindexsuffix(queryparams))).getcstr();
|
---|
98 |
|
---|
99 | // set default stem method from values originally set on prefs page
|
---|
100 | int defaultStemMethod = 0;
|
---|
101 | if (queryparams.casefolding) {
|
---|
102 | defaultStemMethod |= 1;
|
---|
103 | }
|
---|
104 | if (queryparams.stemming) {
|
---|
105 | defaultStemMethod |= 2;
|
---|
106 | }
|
---|
107 |
|
---|
108 | // set default Boolean combiner from all/some setting
|
---|
109 | // if match_mode == 1, ie all, default=1 ie AND
|
---|
110 | // if match_mode == 0, ie some, default=0, ie OR
|
---|
111 | int defaultBoolCombine = 0;
|
---|
112 | if (queryparams.match_mode){
|
---|
113 | defaultBoolCombine = 1;
|
---|
114 | }
|
---|
115 | char* utf8querystring=to_utf8(queryparams.querystring).getcstr();
|
---|
116 | cerr << "**** query string = " << utf8querystring << endl;
|
---|
117 | cerr << "***** index name = " << indexname << endl;
|
---|
118 |
|
---|
119 | text_t cmd = "lucene_query.pl ";
|
---|
120 | cmd += indexname + (text_t)" \"" + to_utf8(queryparams.querystring) + (text_t)"\"";
|
---|
121 |
|
---|
122 | #ifdef __WIN32__
|
---|
123 | FILE *PIN = _popen(cmd.getcstr(), "r");
|
---|
124 | #else
|
---|
125 | FILE *PIN = popen(cmd.getcstr(), "r");
|
---|
126 | #endif
|
---|
127 | if (PIN==NULL) {
|
---|
128 | cerr << "Error: unable to open pipe to " << cmd << endl;
|
---|
129 | return false;
|
---|
130 | }
|
---|
131 |
|
---|
132 | text_t xml_text = "";
|
---|
133 |
|
---|
134 | while (!feof(PIN)) {
|
---|
135 | char buffer[256];
|
---|
136 | int num_bytes = fread(buffer,1,256,PIN);
|
---|
137 | xml_text.appendcarr(buffer,num_bytes);
|
---|
138 | }
|
---|
139 |
|
---|
140 | expat_resultset(xml_text,queryresult);
|
---|
141 |
|
---|
142 | #ifdef __WIN32__
|
---|
143 | _pclose(PIN);
|
---|
144 | #else
|
---|
145 | pclose(PIN);
|
---|
146 | #endif
|
---|
147 |
|
---|
148 | return true;
|
---|
149 |
|
---|
150 | /*
|
---|
151 | // use default query info settings - change to reflect user preferences??
|
---|
152 | QueryInfo queryInfo;
|
---|
153 |
|
---|
154 | SetCStr (queryInfo.docLevel, (queryparams.level.getcstr()));
|
---|
155 | queryInfo.maxDocs = (unsigned long)queryparams.maxdocs;
|
---|
156 | queryInfo.sortByRank = (queryparams.search_type == 1);
|
---|
157 | queryInfo.exactWeights = false;
|
---|
158 | queryInfo.needRankInfo = true; // used for overall term freq as well as ranking
|
---|
159 | queryInfo.needTermFreqs = true;
|
---|
160 |
|
---|
161 | ExtQueryResult queryResult;
|
---|
162 |
|
---|
163 | UCArray queryArray;
|
---|
164 | // greenstone gives us the query encoded in unicode. We want utf8.
|
---|
165 | char* utf8querystring=to_utf8(queryparams.querystring).getcstr();
|
---|
166 | SetCStr(queryArray, utf8querystring);
|
---|
167 | delete utf8querystring;
|
---|
168 |
|
---|
169 | UCArray level;
|
---|
170 | UCArrayClear(level);
|
---|
171 |
|
---|
172 | //set the level for results
|
---|
173 | SetCStr(level, gdbm_level.getcstr());
|
---|
174 |
|
---|
175 |
|
---|
176 | // do the query
|
---|
177 | // LuceneQuery(*indexData, queryInfo, queryTree, queryResult, level); // ****
|
---|
178 |
|
---|
179 |
|
---|
180 | // convert ExtQueryResult to queryresultclass
|
---|
181 |
|
---|
182 | queryresult.docs_matched = (int)queryResult.docs.size();
|
---|
183 |
|
---|
184 | if (queryresult.docs_matched == (int)queryResult.actualNumDocs) {
|
---|
185 | queryresult.is_approx = Exact;
|
---|
186 | }
|
---|
187 | else if (queryresult.docs_matched < (int)queryResult.actualNumDocs) {
|
---|
188 | queryresult.is_approx = MoreThan;
|
---|
189 | }
|
---|
190 | else {
|
---|
191 | queryresult.is_approx = Approximate;
|
---|
192 | }
|
---|
193 |
|
---|
194 | docresultclass doc;
|
---|
195 | for (int i=0; i<(int)queryResult.docs.size(); i++) {
|
---|
196 | doc.clear();
|
---|
197 | doc.docnum = (int)queryResult.levels[i];
|
---|
198 | doc.docweight = queryResult.ranks[i];
|
---|
199 | queryresult.docs.docset[doc.docnum] = doc;
|
---|
200 | queryresult.docs.docorder.push_back(doc.docnum);
|
---|
201 |
|
---|
202 | }
|
---|
203 |
|
---|
204 | // term info
|
---|
205 | termfreqclass term;
|
---|
206 | for (int k=0; k<(int)queryResult.termFreqs.size(); k++) {
|
---|
207 | term.clear();
|
---|
208 | char* termfreq_cstr=GetCStr(queryResult.termFreqs[k].term);
|
---|
209 | term.termstr = to_uni(termfreq_cstr);
|
---|
210 | delete termfreq_cstr;
|
---|
211 | term.termstemstr = term.termstr;
|
---|
212 | // we don't set term.utf8equivterms ?? - jrm21
|
---|
213 | term.termfreq = queryResult.termFreqs[k].termFreq;
|
---|
214 | queryresult.terms.push_back(term);
|
---|
215 | queryresult.orgterms.push_back(term); // should this change??
|
---|
216 |
|
---|
217 | for (int j=0; j<(int)queryResult.termFreqs[k].equivTerms.size(); j++) {
|
---|
218 | char* equivterm_cstr=GetCStr(queryResult.termFreqs[k].equivTerms[j]);
|
---|
219 | queryresult.termvariants.insert(to_uni(equivterm_cstr));
|
---|
220 | delete equivterm_cstr;
|
---|
221 | }
|
---|
222 |
|
---|
223 | }
|
---|
224 | // clean up
|
---|
225 | delete indexname;
|
---|
226 | return true;
|
---|
227 | */
|
---|
228 |
|
---|
229 | return false;
|
---|
230 |
|
---|
231 | }
|
---|
232 |
|
---|
233 |
|
---|
234 | bool lucenesearchclass::browse_search(const queryparamclass &queryparams,
|
---|
235 | int start, int numDocs,
|
---|
236 | queryresultsclass &queryresult) {
|
---|
237 |
|
---|
238 | cerr << "**** Not sure what this function does!" << endl;
|
---|
239 |
|
---|
240 | /*
|
---|
241 | #ifdef __WIN32__
|
---|
242 | char basepath[]="";
|
---|
243 | #else
|
---|
244 | char basepath[] = "/";
|
---|
245 | #endif
|
---|
246 |
|
---|
247 | char *indexname = (filename_cat(collectdir, getindexsuffix(queryparams))).getcstr();
|
---|
248 |
|
---|
249 | UCArray level;
|
---|
250 | UCArrayClear(level);
|
---|
251 |
|
---|
252 | //browse always at top level
|
---|
253 | SetCStr(level, "Doc"); // this name may change.
|
---|
254 |
|
---|
255 |
|
---|
256 | BrowseQueryNode browseNode;
|
---|
257 | browseNode.startPosition = start;
|
---|
258 | browseNode.numTerms = numDocs;
|
---|
259 |
|
---|
260 | BrowseQueryResult browseResult;
|
---|
261 |
|
---|
262 |
|
---|
263 | UCArrayClear(browseNode.term);
|
---|
264 | // greenstone gives us the query encoded in unicode. We want utf8.
|
---|
265 | char* utf8querystring=to_utf8(queryparams.querystring).getcstr();
|
---|
266 | SetCStr(browseNode.term, utf8querystring);
|
---|
267 | delete utf8querystring;
|
---|
268 |
|
---|
269 | // do the actual query
|
---|
270 | // LuceneBrowseQuery(*indexData, level, browseNode, browseResult); // ****
|
---|
271 |
|
---|
272 | // load results into term info
|
---|
273 | termfreqclass term;
|
---|
274 | for (int i=0; i<(int)browseResult.termFreqs.size(); i++) {
|
---|
275 | term.clear();
|
---|
276 | char* term_cstr = GetCStr(browseResult.termFreqs[i].term);
|
---|
277 | term.termstr = to_uni(term_cstr);
|
---|
278 | delete term_cstr;
|
---|
279 | term.termstemstr = term.termstr;
|
---|
280 | term.termfreq = browseResult.termFreqs[i].termFreq;
|
---|
281 | queryresult.terms.push_back(term);
|
---|
282 | queryresult.orgterms.push_back(term);
|
---|
283 |
|
---|
284 | }
|
---|
285 | // clean up
|
---|
286 | delete indexname;
|
---|
287 |
|
---|
288 | return true;
|
---|
289 |
|
---|
290 | */
|
---|
291 |
|
---|
292 | return false;
|
---|
293 | }
|
---|
294 |
|
---|
295 | // the document text for 'docnum' is placed in 'output'
|
---|
296 | // docTargetDocument returns 'true' if it was able to
|
---|
297 | // try to get a document
|
---|
298 | // collection is needed to see if an index from the
|
---|
299 | // collection is loaded. THe default index bits are just there cos
|
---|
300 | // the mg version needs them
|
---|
301 |
|
---|
302 | bool lucenesearchclass::docTargetDocument(const text_t &/*defaultindex*/,
|
---|
303 | const text_t &/*defaultsubcollection*/,
|
---|
304 | const text_t &/*defaultlanguage*/,
|
---|
305 | const text_t &collection,
|
---|
306 | int docnum,
|
---|
307 | text_t &output) {
|
---|
308 |
|
---|
309 | cerr << "**** Should return document text here!" << endl;
|
---|
310 |
|
---|
311 | /*
|
---|
312 | #ifdef __WIN32__
|
---|
313 | char basepath[]="";
|
---|
314 | #else
|
---|
315 | char basepath[] = "/";
|
---|
316 | #endif
|
---|
317 | char *textname = (filename_cat(collectdir, "index", "text", collection)).getcstr();;
|
---|
318 |
|
---|
319 |
|
---|
320 | TextData textdata;
|
---|
321 | if(!textdata.LoadData(basepath, textname)) {
|
---|
322 | cout<<"couldn't load text data\n"<<endl;
|
---|
323 | return false;
|
---|
324 | }
|
---|
325 |
|
---|
326 | UCArray doctext;
|
---|
327 | UCArray level;
|
---|
328 | SetCStr(level, gdbm_level.getcstr());
|
---|
329 | if (!GetDocText(textdata, level, (unsigned long)docnum, doctext)) {
|
---|
330 | cout<<"couldn't retrieve document text\n";
|
---|
331 | return false;
|
---|
332 | }
|
---|
333 |
|
---|
334 | // convert UCArray to text_t
|
---|
335 | output.clear();
|
---|
336 | char* doctext_cstr = GetCStr(doctext);
|
---|
337 | output = to_uni(doctext_cstr); // convert from utf-8 to unicode
|
---|
338 | delete doctext_cstr;
|
---|
339 |
|
---|
340 | // here need to remove the <Document>, <Section>, <Paragraph> tags
|
---|
341 |
|
---|
342 |
|
---|
343 | //clean up
|
---|
344 | textdata.UnloadData ();
|
---|
345 | delete textname;
|
---|
346 |
|
---|
347 | return true;
|
---|
348 |
|
---|
349 | */
|
---|
350 |
|
---|
351 | return false;
|
---|
352 | }
|
---|
353 |
|
---|
354 | // used to clear any cached databases for persistent versions of
|
---|
355 | // Greenstone like the Windows local library
|
---|
356 | void lucenesearchclass::unload_database () {
|
---|
357 | }
|
---|
358 |
|
---|
359 |
|
---|
360 |
|
---|
361 |
|
---|
362 |
|
---|