1 | /**********************************************************************
|
---|
2 | *
|
---|
3 | * lucenequeryfilter.cpp --
|
---|
4 | * Copyright (C) 1999 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 |
|
---|
28 | #include "lucenequeryfilter.h"
|
---|
29 | #include "fileutil.h"
|
---|
30 | #include <assert.h>
|
---|
31 | #include "lucenesearch.h"
|
---|
32 |
|
---|
33 | /////////////////////////////////
|
---|
34 | // functions for queryfilterclass
|
---|
35 | /////////////////////////////////
|
---|
36 |
|
---|
37 |
|
---|
38 | lucenequeryfilterclass::lucenequeryfilterclass ()
|
---|
39 | : queryfilterclass() {
|
---|
40 |
|
---|
41 |
|
---|
42 | FilterOption_t filtopt;
|
---|
43 |
|
---|
44 | // -- onePerTerm Level enumerated
|
---|
45 | // likely to be Doc, Sec, Para, but we dont assume anything now
|
---|
46 | filtopt.clear();
|
---|
47 | filtopt.name = "Level";
|
---|
48 | filtopt.type = FilterOption_t::enumeratedt;
|
---|
49 | filtopt.repeatable = FilterOption_t::onePerTerm;
|
---|
50 | filterOptions["Level"] = filtopt;
|
---|
51 |
|
---|
52 | // -- IndexField, enumerated, used to list available fields
|
---|
53 | filtopt.clear();
|
---|
54 | filtopt.name = "IndexField";
|
---|
55 | filtopt.type = FilterOption_t::enumeratedt;
|
---|
56 | filtopt.repeatable = FilterOption_t::onePerTerm;
|
---|
57 | filtopt.defaultValue = "";
|
---|
58 | filterOptions["IndexField"] = filtopt;
|
---|
59 |
|
---|
60 | }
|
---|
61 |
|
---|
62 | lucenequeryfilterclass::~lucenequeryfilterclass () {
|
---|
63 | }
|
---|
64 |
|
---|
65 |
|
---|
66 | //whether a query is a full text browse
|
---|
67 | bool lucenequeryfilterclass::full_text_browse (int filterRequestOptions) {
|
---|
68 | return (filterRequestOptions & FRfullTextBrowse);
|
---|
69 | }
|
---|
70 |
|
---|
71 | void lucenequeryfilterclass::configure (const text_t &key, const text_tarray &cfgline) {
|
---|
72 | queryfilterclass::configure(key, cfgline);
|
---|
73 |
|
---|
74 | if (key == "indexfieldmap") {
|
---|
75 | indexfieldmap.importmap (cfgline);
|
---|
76 |
|
---|
77 | // update the list of indexes in the filter information
|
---|
78 | text_tarray options;
|
---|
79 | indexfieldmap.gettoarray (options);
|
---|
80 |
|
---|
81 | text_tarray::const_iterator here = options.begin();
|
---|
82 | text_tarray::const_iterator end = options.end();
|
---|
83 | bool start = true;
|
---|
84 | while (here !=end) {
|
---|
85 | if (!(*here).empty()) {
|
---|
86 | filterOptions["IndexField"].validValues.push_back(*here);
|
---|
87 | if (start) {
|
---|
88 | filterOptions["IndexField"].defaultValue = *here;
|
---|
89 | start = false;
|
---|
90 | }
|
---|
91 | }
|
---|
92 | ++here;
|
---|
93 | }
|
---|
94 | } else if (key == "indexlevels") {
|
---|
95 | text_tarray::const_iterator here = cfgline.begin();
|
---|
96 | text_tarray::const_iterator end = cfgline.end();
|
---|
97 | bool first=true;
|
---|
98 | while (here != end) {
|
---|
99 | if (!(*here).empty()) {
|
---|
100 | if (first) {
|
---|
101 | first = false;
|
---|
102 | // the default is the first value
|
---|
103 | filterOptions["Level"].defaultValue = *here;
|
---|
104 | }
|
---|
105 | filterOptions["Level"].validValues.push_back(*here);
|
---|
106 | }
|
---|
107 | ++here;
|
---|
108 | }
|
---|
109 | } else if (key == "textlevel") {
|
---|
110 | ((lucenesearchclass *)textsearchptr)->set_gdbm_level( cfgline[0]);
|
---|
111 | }
|
---|
112 |
|
---|
113 | }
|
---|
114 |
|
---|
115 |
|
---|
116 | void lucenequeryfilterclass::filter(const FilterRequest_t &request,
|
---|
117 | FilterResponse_t &response,
|
---|
118 | comerror_t &err, ostream &logout) {
|
---|
119 |
|
---|
120 | outconvertclass text_t2ascii;
|
---|
121 |
|
---|
122 | response.clear ();
|
---|
123 | err = noError;
|
---|
124 | if (gdbmptr == NULL) {
|
---|
125 | // most likely a configuration problem
|
---|
126 | logout << text_t2ascii
|
---|
127 | << "configuration error: queryfilter contains a null gdbmclass\n\n";
|
---|
128 | err = configurationError;
|
---|
129 | return;
|
---|
130 | }
|
---|
131 | if (textsearchptr == NULL) {
|
---|
132 | // most likely a configuration problem
|
---|
133 | logout << text_t2ascii
|
---|
134 | << "configuration error: queryfilter contains a null textsearchclass (lucene)\n\n";
|
---|
135 | err = configurationError;
|
---|
136 | return;
|
---|
137 | }
|
---|
138 | if (full_text_browse(request.filterResultOptions)) {
|
---|
139 | browsefilter(request, response, err, logout);
|
---|
140 | return;
|
---|
141 | }
|
---|
142 | // open the database
|
---|
143 | gdbmptr->setlogout(&logout);
|
---|
144 | if (!gdbmptr->opendatabase (gdbm_filename, GDBM_READER, 100, false)) {
|
---|
145 | // most likely a system problem (we have already checked that the
|
---|
146 | // gdbm database exists)
|
---|
147 | logout << text_t2ascii
|
---|
148 | << "system problem: open on gdbm database \""
|
---|
149 | << gdbm_filename << "\" failed\n\n";
|
---|
150 | err = systemProblem;
|
---|
151 | return;
|
---|
152 | }
|
---|
153 |
|
---|
154 |
|
---|
155 | // get the query parameters
|
---|
156 | int startresults, endresults;
|
---|
157 | text_t phrasematch; // not used here any more
|
---|
158 | vector<queryparamclass> queryfilterparams;
|
---|
159 | parse_query_params (request, queryfilterparams, startresults,
|
---|
160 | endresults, phrasematch, logout);
|
---|
161 |
|
---|
162 |
|
---|
163 | // do query
|
---|
164 | queryresultsclass queryresults;
|
---|
165 | do_multi_query (request, queryfilterparams, queryresults, err, logout);
|
---|
166 | if (err != noError) return;
|
---|
167 | // assemble document results
|
---|
168 | if (need_matching_docs (request.filterResultOptions)) {
|
---|
169 |
|
---|
170 | int resultnum = 1;
|
---|
171 | ResultDocInfo_t resultdoc;
|
---|
172 | text_t trans_OID;
|
---|
173 | vector<int>::iterator docorder_here = queryresults.docs.docorder.begin();
|
---|
174 | vector<int>::iterator docorder_end = queryresults.docs.docorder.end();
|
---|
175 |
|
---|
176 | if (endresults == -1) endresults = MAXNUMDOCS;
|
---|
177 | while (docorder_here != docorder_end) {
|
---|
178 | if (resultnum > endresults) break;
|
---|
179 |
|
---|
180 | // translate the document number
|
---|
181 | if (!translate(gdbmptr, *docorder_here, trans_OID)) {
|
---|
182 | logout << text_t2ascii
|
---|
183 | << "warning: could not translate lucene document number \""
|
---|
184 | << *docorder_here << "\" to OID.\n\n";
|
---|
185 |
|
---|
186 | } else {
|
---|
187 | docresultmap::iterator docset_here = queryresults.docs.docset.find (*docorder_here);
|
---|
188 |
|
---|
189 | // see if there is a result for this number,
|
---|
190 | // if it is in the request set (or the request set is empty)
|
---|
191 | if (docset_here != queryresults.docs.docset.end() &&
|
---|
192 | (request.docSet.empty() || in_set(request.docSet, trans_OID))) {
|
---|
193 | if (resultnum >= startresults) {
|
---|
194 | // add this document
|
---|
195 | resultdoc.OID = trans_OID;
|
---|
196 | resultdoc.result_num = resultnum;
|
---|
197 | resultdoc.ranking = (int)((*docset_here).second.docweight * 10000.0 + 0.5);
|
---|
198 |
|
---|
199 | response.docInfo.push_back (resultdoc);
|
---|
200 | }
|
---|
201 |
|
---|
202 | ++resultnum;
|
---|
203 | }
|
---|
204 | } // else
|
---|
205 |
|
---|
206 | ++docorder_here;
|
---|
207 | }
|
---|
208 | } // if need matching docs
|
---|
209 |
|
---|
210 | // assemble the term results
|
---|
211 | if (need_term_info(request.filterResultOptions)) {
|
---|
212 | // note: the terms have already been sorted and uniqued - ?? have they??
|
---|
213 |
|
---|
214 | TermInfo_t terminfo;
|
---|
215 | bool terms_first = true;
|
---|
216 |
|
---|
217 | termfreqclassarray::iterator terms_here = queryresults.terms.begin();
|
---|
218 | termfreqclassarray::iterator terms_end = queryresults.terms.end();
|
---|
219 |
|
---|
220 | while (terms_here != terms_end) {
|
---|
221 | terminfo.clear();
|
---|
222 | terminfo.term = (*terms_here).termstr;
|
---|
223 | terminfo.freq = (*terms_here).termfreq;
|
---|
224 |
|
---|
225 | // this bit gets the matchTerms ie the equivalent (stem/casefold) terms
|
---|
226 | if (terms_first) {
|
---|
227 | text_tset::iterator termvariants_here = queryresults.termvariants.begin();
|
---|
228 | text_tset::iterator termvariants_end = queryresults.termvariants.end();
|
---|
229 | while (termvariants_here != termvariants_end) {
|
---|
230 | terminfo.matchTerms.push_back (*termvariants_here);
|
---|
231 | ++termvariants_here;
|
---|
232 | }
|
---|
233 | }
|
---|
234 | terms_first = false;
|
---|
235 |
|
---|
236 | response.termInfo.push_back (terminfo);
|
---|
237 |
|
---|
238 | ++terms_here;
|
---|
239 | }
|
---|
240 |
|
---|
241 | // add the stop words
|
---|
242 | text_tset::iterator stopwords_here = queryresults.stopwords.begin();
|
---|
243 | text_tset::iterator stopwords_end = queryresults.stopwords.end();
|
---|
244 | while (stopwords_here != stopwords_end) {
|
---|
245 | response.stopwords.insert(*stopwords_here);
|
---|
246 | ++stopwords_here;
|
---|
247 | }
|
---|
248 | }
|
---|
249 |
|
---|
250 | response.numDocs = queryresults.docs_matched;
|
---|
251 | response.isApprox = queryresults.is_approx;
|
---|
252 | }
|
---|
253 |
|
---|
254 | void lucenequeryfilterclass::browsefilter(const FilterRequest_t &request,
|
---|
255 | FilterResponse_t &response,
|
---|
256 | comerror_t &err, ostream &logout) {
|
---|
257 |
|
---|
258 | outconvertclass text_t2ascii;
|
---|
259 |
|
---|
260 | // get the query parameters
|
---|
261 | int startresults, endresults;
|
---|
262 | text_t phrasematch; // not used here any more, just have it so can use
|
---|
263 | // parse_query_params function
|
---|
264 |
|
---|
265 | vector<queryparamclass> queryfilterparams;
|
---|
266 | parse_query_params (request, queryfilterparams, startresults,
|
---|
267 | endresults, phrasematch, logout);
|
---|
268 |
|
---|
269 | vector<queryparamclass>::const_iterator query_here = queryfilterparams.begin();
|
---|
270 |
|
---|
271 | // do query
|
---|
272 | queryresultsclass queryresults;
|
---|
273 | queryresults.clear();
|
---|
274 |
|
---|
275 | int numDocs = endresults-startresults;
|
---|
276 | textsearchptr->setcollectdir (collectdir);
|
---|
277 |
|
---|
278 | if (!((lucenesearchclass*)textsearchptr)->browse_search((*query_here), startresults, numDocs, queryresults)) {
|
---|
279 | // most likely a system problem
|
---|
280 | logout << text_t2ascii
|
---|
281 | << "system problem: could not do full text browse with lucene for index \""
|
---|
282 | << (*query_here).index << (*query_here).subcollection
|
---|
283 | << (*query_here).language << "\".\n\n";
|
---|
284 | err = systemProblem;
|
---|
285 | return;
|
---|
286 | }
|
---|
287 |
|
---|
288 | // assemble the term results
|
---|
289 | TermInfo_t terminfo;
|
---|
290 |
|
---|
291 | termfreqclassarray::iterator terms_here = queryresults.terms.begin();
|
---|
292 | termfreqclassarray::iterator terms_end = queryresults.terms.end();
|
---|
293 |
|
---|
294 | while (terms_here != terms_end) {
|
---|
295 | terminfo.clear();
|
---|
296 | terminfo.term = (*terms_here).termstr;
|
---|
297 | terminfo.freq = (*terms_here).termfreq;
|
---|
298 |
|
---|
299 | response.termInfo.push_back (terminfo);
|
---|
300 |
|
---|
301 | ++terms_here;
|
---|
302 | }
|
---|
303 |
|
---|
304 |
|
---|
305 | }
|
---|
306 |
|
---|
307 | // lucenesearchptr and gdbmptr are assumed to be valid
|
---|
308 | void lucenequeryfilterclass::do_multi_query (const FilterRequest_t &request,
|
---|
309 | const vector<queryparamclass> &query_params,
|
---|
310 | queryresultsclass &multiresults,
|
---|
311 | comerror_t &err, ostream &logout) {
|
---|
312 | outconvertclass text_t2ascii;
|
---|
313 |
|
---|
314 | err = noError;
|
---|
315 | textsearchptr->setcollectdir (collectdir);
|
---|
316 | multiresults.clear();
|
---|
317 |
|
---|
318 | vector<queryparamclass>::const_iterator query_here = query_params.begin();
|
---|
319 | vector<queryparamclass>::const_iterator query_end = query_params.end();
|
---|
320 | while (query_here != query_end) {
|
---|
321 | queryresultsclass thisqueryresults;
|
---|
322 | if (!textsearchptr->search((*query_here), thisqueryresults)) {
|
---|
323 | // most likely a system problem
|
---|
324 | logout << text_t2ascii
|
---|
325 | << "system problem: could not do search with lucene for index \""
|
---|
326 | << (*query_here).index << (*query_here).level
|
---|
327 | << (*query_here).subcollection
|
---|
328 | << (*query_here).language << "\".\n\n";
|
---|
329 | err = systemProblem;
|
---|
330 | return;
|
---|
331 | }
|
---|
332 |
|
---|
333 | // check for syntax error
|
---|
334 | if (thisqueryresults.syntax_error==true) {
|
---|
335 | logout << text_t2ascii
|
---|
336 | << "syntax problem: invalid query string \""
|
---|
337 | << (*query_here).querystring<<"\".\n";
|
---|
338 | err = syntaxError;
|
---|
339 | return;
|
---|
340 | }
|
---|
341 | // combine the results
|
---|
342 | if (need_matching_docs (request.filterResultOptions)) {
|
---|
343 |
|
---|
344 | if (query_params.size() == 1) {
|
---|
345 | multiresults.docs = thisqueryresults.docs; // just one set of results
|
---|
346 | multiresults.docs_matched = thisqueryresults.docs_matched;
|
---|
347 | multiresults.is_approx = thisqueryresults.is_approx;
|
---|
348 |
|
---|
349 | } else {
|
---|
350 | if ((*query_here).combinequery == "and") {
|
---|
351 | multiresults.docs.combine_and (thisqueryresults.docs);
|
---|
352 | } else if ((*query_here).combinequery == "or") {
|
---|
353 | multiresults.docs.combine_or (thisqueryresults.docs);
|
---|
354 | } else if ((*query_here).combinequery == "not") {
|
---|
355 | multiresults.docs.combine_not (thisqueryresults.docs);
|
---|
356 | }
|
---|
357 | multiresults.docs_matched = multiresults.docs.docset.size();
|
---|
358 | multiresults.is_approx = Exact;
|
---|
359 | }
|
---|
360 | }
|
---|
361 |
|
---|
362 | // combine the term information
|
---|
363 | if (need_term_info (request.filterResultOptions)) {
|
---|
364 | // append the terms
|
---|
365 | multiresults.orgterms.insert(multiresults.orgterms.end(),
|
---|
366 | thisqueryresults.orgterms.begin(),
|
---|
367 | thisqueryresults.orgterms.end());
|
---|
368 |
|
---|
369 |
|
---|
370 | // add the term variants -
|
---|
371 | text_tset::iterator termvar_here = thisqueryresults.termvariants.begin();
|
---|
372 | text_tset::iterator termvar_end = thisqueryresults.termvariants.end();
|
---|
373 | while (termvar_here != termvar_end) {
|
---|
374 | multiresults.termvariants.insert(*termvar_here);
|
---|
375 | ++termvar_here;
|
---|
376 | }
|
---|
377 |
|
---|
378 | // add the stop words
|
---|
379 | text_tset::iterator stopwords_here = thisqueryresults.stopwords.begin();
|
---|
380 | text_tset::iterator stopwords_end = thisqueryresults.stopwords.end();
|
---|
381 | while (stopwords_here != stopwords_end) {
|
---|
382 | multiresults.stopwords.insert(*stopwords_here);
|
---|
383 | ++stopwords_here;
|
---|
384 | }
|
---|
385 | }
|
---|
386 |
|
---|
387 | ++query_here;
|
---|
388 | }
|
---|
389 |
|
---|
390 | // sort and unique the query terms
|
---|
391 | multiresults.sortuniqqueryterms ();
|
---|
392 | }
|
---|
393 |
|
---|
394 |
|
---|
395 |
|
---|