[8027] | 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 | #include "lucenequeryfilter.h"
|
---|
| 27 | #include "fileutil.h"
|
---|
| 28 | #include "lucenesearch.h"
|
---|
| 29 |
|
---|
| 30 | /////////////////////////////////
|
---|
| 31 | // functions for queryfilterclass
|
---|
| 32 | /////////////////////////////////
|
---|
| 33 |
|
---|
| 34 |
|
---|
| 35 | lucenequeryfilterclass::lucenequeryfilterclass ()
|
---|
| 36 | : queryfilterclass() {
|
---|
| 37 |
|
---|
| 38 |
|
---|
| 39 | FilterOption_t filtopt;
|
---|
| 40 |
|
---|
| 41 | // -- onePerTerm Level enumerated
|
---|
| 42 | // likely to be Doc, Sec, Para, but we dont assume anything now
|
---|
| 43 | filtopt.clear();
|
---|
| 44 | filtopt.name = "Level";
|
---|
| 45 | filtopt.type = FilterOption_t::enumeratedt;
|
---|
| 46 | filtopt.repeatable = FilterOption_t::onePerTerm;
|
---|
[20727] | 47 | filtopt.defaultValue = "";
|
---|
[8027] | 48 | filterOptions["Level"] = filtopt;
|
---|
| 49 |
|
---|
| 50 | // -- IndexField, enumerated, used to list available fields
|
---|
| 51 | filtopt.clear();
|
---|
| 52 | filtopt.name = "IndexField";
|
---|
| 53 | filtopt.type = FilterOption_t::enumeratedt;
|
---|
| 54 | filtopt.repeatable = FilterOption_t::onePerTerm;
|
---|
| 55 | filtopt.defaultValue = "";
|
---|
| 56 | filterOptions["IndexField"] = filtopt;
|
---|
| 57 |
|
---|
| 58 | }
|
---|
| 59 |
|
---|
| 60 | lucenequeryfilterclass::~lucenequeryfilterclass () {
|
---|
| 61 | }
|
---|
| 62 |
|
---|
| 63 |
|
---|
| 64 | //whether a query is a full text browse
|
---|
| 65 | bool lucenequeryfilterclass::full_text_browse (int filterRequestOptions) {
|
---|
| 66 | return (filterRequestOptions & FRfullTextBrowse);
|
---|
| 67 | }
|
---|
| 68 |
|
---|
| 69 | void lucenequeryfilterclass::configure (const text_t &key, const text_tarray &cfgline) {
|
---|
| 70 | queryfilterclass::configure(key, cfgline);
|
---|
| 71 |
|
---|
| 72 | if (key == "indexfieldmap") {
|
---|
| 73 | indexfieldmap.importmap (cfgline);
|
---|
| 74 |
|
---|
| 75 | // update the list of indexes in the filter information
|
---|
| 76 | text_tarray options;
|
---|
| 77 | indexfieldmap.gettoarray (options);
|
---|
[20727] | 78 | filterOptions["IndexField"].validValues = options;
|
---|
| 79 | } else if (key == "levelmap") {
|
---|
| 80 | levelmap.importmap (cfgline);
|
---|
[8027] | 81 | } else if (key == "indexlevels") {
|
---|
| 82 | text_tarray::const_iterator here = cfgline.begin();
|
---|
| 83 | text_tarray::const_iterator end = cfgline.end();
|
---|
| 84 | while (here != end) {
|
---|
| 85 | if (!(*here).empty()) {
|
---|
| 86 | filterOptions["Level"].validValues.push_back(*here);
|
---|
| 87 | }
|
---|
[9620] | 88 | ++here;
|
---|
[8027] | 89 | }
|
---|
| 90 | } else if (key == "textlevel") {
|
---|
[15594] | 91 | ((lucenesearchclass *)textsearchptr)->set_text_level(cfgline[0]);
|
---|
[20727] | 92 | } else if (key == "defaultindex") {
|
---|
| 93 | indexfieldmap.from2to (cfgline[0], filterOptions["IndexField"].defaultValue);
|
---|
| 94 | } else if (key == "defaultlevel") {
|
---|
| 95 | levelmap.from2to (cfgline[0], filterOptions["Level"].defaultValue);
|
---|
[8027] | 96 | }
|
---|
| 97 |
|
---|
| 98 | }
|
---|
| 99 |
|
---|
[20727] | 100 | bool lucenequeryfilterclass::init (ostream &logout) {
|
---|
| 101 |
|
---|
| 102 | if (!queryfilterclass::init(logout)) {
|
---|
| 103 | return false;
|
---|
| 104 | }
|
---|
| 105 |
|
---|
| 106 | if (filterOptions["IndexField"].defaultValue.empty()) {
|
---|
| 107 | // use first index in map as default if no default is set explicitly
|
---|
| 108 | text_tarray fromarray;
|
---|
| 109 | indexfieldmap.getfromarray(fromarray);
|
---|
| 110 | if (fromarray.size()) {
|
---|
| 111 | filterOptions["IndexField"].defaultValue = fromarray[0];
|
---|
| 112 | }
|
---|
| 113 | }
|
---|
| 114 | if (filterOptions["Levels"].defaultValue.empty()) {
|
---|
| 115 | // use first level as default if no default is set explicitly
|
---|
| 116 | if (!filterOptions["Level"].validValues[0].empty())
|
---|
| 117 | filterOptions["Levels"].defaultValue = filterOptions["Level"].validValues[0];
|
---|
| 118 | }
|
---|
[8027] | 119 |
|
---|
[20727] | 120 | return true;
|
---|
| 121 | }
|
---|
| 122 |
|
---|
[8027] | 123 | void lucenequeryfilterclass::filter(const FilterRequest_t &request,
|
---|
| 124 | FilterResponse_t &response,
|
---|
| 125 | comerror_t &err, ostream &logout) {
|
---|
| 126 |
|
---|
| 127 | outconvertclass text_t2ascii;
|
---|
| 128 |
|
---|
| 129 | response.clear ();
|
---|
| 130 | err = noError;
|
---|
[15558] | 131 | if (db_ptr == NULL) {
|
---|
[8027] | 132 | // most likely a configuration problem
|
---|
| 133 | logout << text_t2ascii
|
---|
[15558] | 134 | << "configuration error: queryfilter contains a null dbclass\n\n";
|
---|
[8027] | 135 | err = configurationError;
|
---|
| 136 | return;
|
---|
| 137 | }
|
---|
| 138 | if (textsearchptr == NULL) {
|
---|
| 139 | // most likely a configuration problem
|
---|
| 140 | logout << text_t2ascii
|
---|
[9110] | 141 | << "configuration error: queryfilter contains a null textsearchclass (lucene)\n\n";
|
---|
[8027] | 142 | err = configurationError;
|
---|
| 143 | return;
|
---|
| 144 | }
|
---|
| 145 | if (full_text_browse(request.filterResultOptions)) {
|
---|
| 146 | browsefilter(request, response, err, logout);
|
---|
| 147 | return;
|
---|
| 148 | }
|
---|
| 149 | // open the database
|
---|
[15558] | 150 | db_ptr->setlogout(&logout);
|
---|
| 151 | if (!db_ptr->opendatabase (db_filename, DB_READER, 100, false)) {
|
---|
| 152 | // most likely a system problem (we have already checked that the database exists)
|
---|
[8027] | 153 | logout << text_t2ascii
|
---|
[15558] | 154 | << "system problem: open on database \"" << db_filename << "\" failed\n\n";
|
---|
[8027] | 155 | err = systemProblem;
|
---|
| 156 | return;
|
---|
| 157 | }
|
---|
| 158 |
|
---|
| 159 |
|
---|
| 160 | // get the query parameters
|
---|
| 161 | int startresults, endresults;
|
---|
| 162 | text_t phrasematch; // not used here any more
|
---|
| 163 | vector<queryparamclass> queryfilterparams;
|
---|
| 164 | parse_query_params (request, queryfilterparams, startresults,
|
---|
| 165 | endresults, phrasematch, logout);
|
---|
| 166 |
|
---|
| 167 |
|
---|
| 168 | // do query
|
---|
| 169 | queryresultsclass queryresults;
|
---|
| 170 | do_multi_query (request, queryfilterparams, queryresults, err, logout);
|
---|
[12421] | 171 | response.error_message = queryresults.error_message;
|
---|
[8027] | 172 | if (err != noError) return;
|
---|
[12421] | 173 |
|
---|
[8027] | 174 | // assemble document results
|
---|
[16947] | 175 | if (need_matching_docs (request.filterResultOptions))
|
---|
| 176 | {
|
---|
| 177 | // Loop through the query results (ordered by ranking)
|
---|
[8027] | 178 | int resultnum = 1;
|
---|
[16947] | 179 | vector<text_t>::iterator docorder_iterator = queryresults.docs.docorder.begin();
|
---|
| 180 | while (docorder_iterator != queryresults.docs.docorder.end())
|
---|
| 181 | {
|
---|
| 182 | text_t doc_OID = (*docorder_iterator);
|
---|
| 183 | // logout << "Matching doc OID: " << doc_OID << endl;
|
---|
[8027] | 184 |
|
---|
[16947] | 185 | // Make sure this result is in the docset, and either in the request set or the request set is empty
|
---|
| 186 | docresultmap::iterator doc_result = queryresults.docs.docset.find (doc_OID);
|
---|
| 187 | if (doc_result != queryresults.docs.docset.end() && (request.docSet.empty() || in_set(request.docSet, doc_OID)))
|
---|
[12655] | 188 | {
|
---|
[16947] | 189 | // Add the matching document
|
---|
| 190 | ResultDocInfo_t resultdoc;
|
---|
| 191 | resultdoc.OID = doc_OID;
|
---|
| 192 | resultdoc.result_num = resultnum;
|
---|
| 193 | resultdoc.ranking = (int)((*doc_result).second.docweight * 10000.0 + 0.5);
|
---|
| 194 | resultdoc.num_terms_matched = (*doc_result).second.num_query_terms_matched;
|
---|
| 195 | response.docInfo.push_back (resultdoc);
|
---|
[8027] | 196 |
|
---|
[16947] | 197 | resultnum++;
|
---|
| 198 | }
|
---|
[8027] | 199 |
|
---|
[16947] | 200 | docorder_iterator++;
|
---|
| 201 | }
|
---|
| 202 | }
|
---|
[12655] | 203 |
|
---|
[8027] | 204 | // assemble the term results
|
---|
| 205 | if (need_term_info(request.filterResultOptions)) {
|
---|
| 206 | // note: the terms have already been sorted and uniqued - ?? have they??
|
---|
| 207 |
|
---|
| 208 | TermInfo_t terminfo;
|
---|
| 209 | bool terms_first = true;
|
---|
| 210 |
|
---|
| 211 | termfreqclassarray::iterator terms_here = queryresults.terms.begin();
|
---|
| 212 | termfreqclassarray::iterator terms_end = queryresults.terms.end();
|
---|
| 213 |
|
---|
| 214 | while (terms_here != terms_end) {
|
---|
| 215 | terminfo.clear();
|
---|
| 216 | terminfo.term = (*terms_here).termstr;
|
---|
| 217 | terminfo.freq = (*terms_here).termfreq;
|
---|
[13063] | 218 | // lucene doesn't return any termvariants at this stage,
|
---|
| 219 | // so make sure the original term is set
|
---|
| 220 | terminfo.matchTerms.push_back(terminfo.term);
|
---|
| 221 |
|
---|
[8027] | 222 | // this bit gets the matchTerms ie the equivalent (stem/casefold) terms
|
---|
| 223 | if (terms_first) {
|
---|
| 224 | text_tset::iterator termvariants_here = queryresults.termvariants.begin();
|
---|
| 225 | text_tset::iterator termvariants_end = queryresults.termvariants.end();
|
---|
| 226 | while (termvariants_here != termvariants_end) {
|
---|
| 227 | terminfo.matchTerms.push_back (*termvariants_here);
|
---|
[9620] | 228 | ++termvariants_here;
|
---|
[8027] | 229 | }
|
---|
| 230 | }
|
---|
| 231 | terms_first = false;
|
---|
| 232 |
|
---|
| 233 | response.termInfo.push_back (terminfo);
|
---|
| 234 |
|
---|
[9620] | 235 | ++terms_here;
|
---|
[8027] | 236 | }
|
---|
[12380] | 237 |
|
---|
| 238 | // add the stop words
|
---|
| 239 | text_tset::iterator stopwords_here = queryresults.stopwords.begin();
|
---|
| 240 | text_tset::iterator stopwords_end = queryresults.stopwords.end();
|
---|
| 241 | while (stopwords_here != stopwords_end) {
|
---|
| 242 | response.stopwords.insert(*stopwords_here);
|
---|
| 243 | ++stopwords_here;
|
---|
| 244 | }
|
---|
[8027] | 245 | }
|
---|
| 246 |
|
---|
[15558] | 247 | db_ptr->closedatabase(); // Important that local library doesn't leave any files open
|
---|
[8027] | 248 | response.numDocs = queryresults.docs_matched;
|
---|
| 249 | response.isApprox = queryresults.is_approx;
|
---|
| 250 | }
|
---|
| 251 |
|
---|
| 252 | void lucenequeryfilterclass::browsefilter(const FilterRequest_t &request,
|
---|
| 253 | FilterResponse_t &response,
|
---|
| 254 | comerror_t &err, ostream &logout) {
|
---|
| 255 |
|
---|
| 256 | outconvertclass text_t2ascii;
|
---|
| 257 |
|
---|
| 258 | // get the query parameters
|
---|
| 259 | int startresults, endresults;
|
---|
| 260 | text_t phrasematch; // not used here any more, just have it so can use
|
---|
| 261 | // parse_query_params function
|
---|
| 262 |
|
---|
| 263 | vector<queryparamclass> queryfilterparams;
|
---|
| 264 | parse_query_params (request, queryfilterparams, startresults,
|
---|
| 265 | endresults, phrasematch, logout);
|
---|
| 266 |
|
---|
[22050] | 267 | vector<queryparamclass>::const_iterator query_here = queryfilterparams.begin();
|
---|
[8027] | 268 |
|
---|
| 269 | // do query
|
---|
| 270 | queryresultsclass queryresults;
|
---|
| 271 | queryresults.clear();
|
---|
| 272 |
|
---|
| 273 | int numDocs = endresults-startresults;
|
---|
| 274 | textsearchptr->setcollectdir (collectdir);
|
---|
| 275 |
|
---|
| 276 | if (!((lucenesearchclass*)textsearchptr)->browse_search((*query_here), startresults, numDocs, queryresults)) {
|
---|
| 277 | // most likely a system problem
|
---|
| 278 | logout << text_t2ascii
|
---|
| 279 | << "system problem: could not do full text browse with lucene for index \""
|
---|
| 280 | << (*query_here).index << (*query_here).subcollection
|
---|
| 281 | << (*query_here).language << "\".\n\n";
|
---|
| 282 | err = systemProblem;
|
---|
| 283 | return;
|
---|
| 284 | }
|
---|
| 285 |
|
---|
| 286 | // assemble the term results
|
---|
| 287 | TermInfo_t terminfo;
|
---|
| 288 |
|
---|
| 289 | termfreqclassarray::iterator terms_here = queryresults.terms.begin();
|
---|
| 290 | termfreqclassarray::iterator terms_end = queryresults.terms.end();
|
---|
| 291 |
|
---|
| 292 | while (terms_here != terms_end) {
|
---|
| 293 | terminfo.clear();
|
---|
| 294 | terminfo.term = (*terms_here).termstr;
|
---|
| 295 | terminfo.freq = (*terms_here).termfreq;
|
---|
| 296 |
|
---|
| 297 | response.termInfo.push_back (terminfo);
|
---|
| 298 |
|
---|
[9620] | 299 | ++terms_here;
|
---|
[8027] | 300 | }
|
---|
| 301 |
|
---|
| 302 |
|
---|
| 303 | }
|
---|
| 304 |
|
---|
[15558] | 305 | // lucenesearchptr and db_ptr are assumed to be valid
|
---|
[8027] | 306 | void lucenequeryfilterclass::do_multi_query (const FilterRequest_t &request,
|
---|
| 307 | const vector<queryparamclass> &query_params,
|
---|
| 308 | queryresultsclass &multiresults,
|
---|
| 309 | comerror_t &err, ostream &logout) {
|
---|
| 310 | outconvertclass text_t2ascii;
|
---|
| 311 |
|
---|
| 312 | err = noError;
|
---|
| 313 | textsearchptr->setcollectdir (collectdir);
|
---|
| 314 | multiresults.clear();
|
---|
| 315 |
|
---|
| 316 | vector<queryparamclass>::const_iterator query_here = query_params.begin();
|
---|
| 317 | vector<queryparamclass>::const_iterator query_end = query_params.end();
|
---|
| 318 | while (query_here != query_end) {
|
---|
| 319 | queryresultsclass thisqueryresults;
|
---|
| 320 | if (!textsearchptr->search((*query_here), thisqueryresults)) {
|
---|
| 321 | // most likely a system problem
|
---|
| 322 | logout << text_t2ascii
|
---|
| 323 | << "system problem: could not do search with lucene for index \""
|
---|
[9090] | 324 | << (*query_here).index << (*query_here).level
|
---|
| 325 | << (*query_here).subcollection
|
---|
[8027] | 326 | << (*query_here).language << "\".\n\n";
|
---|
| 327 | err = systemProblem;
|
---|
| 328 | return;
|
---|
| 329 | }
|
---|
| 330 |
|
---|
| 331 | // check for syntax error
|
---|
| 332 | if (thisqueryresults.syntax_error==true) {
|
---|
| 333 | logout << text_t2ascii
|
---|
| 334 | << "syntax problem: invalid query string \""
|
---|
| 335 | << (*query_here).querystring<<"\".\n";
|
---|
| 336 | err = syntaxError;
|
---|
| 337 | return;
|
---|
| 338 | }
|
---|
| 339 | // combine the results
|
---|
| 340 | if (need_matching_docs (request.filterResultOptions)) {
|
---|
| 341 |
|
---|
| 342 | if (query_params.size() == 1) {
|
---|
[12421] | 343 | multiresults.error_message = thisqueryresults.error_message;
|
---|
[8027] | 344 | multiresults.docs = thisqueryresults.docs; // just one set of results
|
---|
| 345 | multiresults.docs_matched = thisqueryresults.docs_matched;
|
---|
| 346 | multiresults.is_approx = thisqueryresults.is_approx;
|
---|
| 347 |
|
---|
| 348 | } else {
|
---|
| 349 | if ((*query_here).combinequery == "and") {
|
---|
| 350 | multiresults.docs.combine_and (thisqueryresults.docs);
|
---|
| 351 | } else if ((*query_here).combinequery == "or") {
|
---|
| 352 | multiresults.docs.combine_or (thisqueryresults.docs);
|
---|
| 353 | } else if ((*query_here).combinequery == "not") {
|
---|
| 354 | multiresults.docs.combine_not (thisqueryresults.docs);
|
---|
| 355 | }
|
---|
| 356 | multiresults.docs_matched = multiresults.docs.docset.size();
|
---|
| 357 | multiresults.is_approx = Exact;
|
---|
| 358 | }
|
---|
| 359 | }
|
---|
| 360 |
|
---|
| 361 | // combine the term information
|
---|
| 362 | if (need_term_info (request.filterResultOptions)) {
|
---|
| 363 | // append the terms
|
---|
| 364 | multiresults.orgterms.insert(multiresults.orgterms.end(),
|
---|
| 365 | thisqueryresults.orgterms.begin(),
|
---|
| 366 | thisqueryresults.orgterms.end());
|
---|
| 367 |
|
---|
| 368 |
|
---|
| 369 | // add the term variants -
|
---|
| 370 | text_tset::iterator termvar_here = thisqueryresults.termvariants.begin();
|
---|
| 371 | text_tset::iterator termvar_end = thisqueryresults.termvariants.end();
|
---|
| 372 | while (termvar_here != termvar_end) {
|
---|
| 373 | multiresults.termvariants.insert(*termvar_here);
|
---|
[9620] | 374 | ++termvar_here;
|
---|
[8027] | 375 | }
|
---|
[12380] | 376 |
|
---|
| 377 | // add the stop words
|
---|
| 378 | text_tset::iterator stopwords_here = thisqueryresults.stopwords.begin();
|
---|
| 379 | text_tset::iterator stopwords_end = thisqueryresults.stopwords.end();
|
---|
| 380 | while (stopwords_here != stopwords_end) {
|
---|
| 381 | multiresults.stopwords.insert(*stopwords_here);
|
---|
| 382 | ++stopwords_here;
|
---|
| 383 | }
|
---|
[8027] | 384 | }
|
---|
| 385 |
|
---|
[9620] | 386 | ++query_here;
|
---|
[8027] | 387 | }
|
---|
| 388 |
|
---|
| 389 | // sort and unique the query terms
|
---|
| 390 | multiresults.sortuniqqueryterms ();
|
---|
| 391 | }
|
---|
| 392 |
|
---|
| 393 |
|
---|
| 394 |
|
---|