| 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 | response.error_message = queryresults.error_message;
|
|---|
| 167 | if (err != noError) return;
|
|---|
| 168 |
|
|---|
| 169 | // assemble document results
|
|---|
| 170 | if (need_matching_docs (request.filterResultOptions)) {
|
|---|
| 171 |
|
|---|
| 172 | int resultnum = 1;
|
|---|
| 173 | ResultDocInfo_t resultdoc;
|
|---|
| 174 | text_t trans_OID;
|
|---|
| 175 | vector<int>::iterator docorder_here = queryresults.docs.docorder.begin();
|
|---|
| 176 | vector<int>::iterator docorder_end = queryresults.docs.docorder.end();
|
|---|
| 177 |
|
|---|
| 178 | // Now handled by Lucene directly
|
|---|
| 179 | //if (endresults == -1) endresults = MAXNUMDOCS;
|
|---|
| 180 |
|
|---|
| 181 | while (docorder_here != docorder_end)
|
|---|
| 182 | {
|
|---|
| 183 | // Now handled by Lucene directly
|
|---|
| 184 | //if (resultnum > endresults) break;
|
|---|
| 185 |
|
|---|
| 186 | // translate the document number
|
|---|
| 187 | if (!translate(gdbmptr, *docorder_here, trans_OID))
|
|---|
| 188 | {
|
|---|
| 189 | logout << text_t2ascii
|
|---|
| 190 | << "warning: could not translate lucene document number \""
|
|---|
| 191 | << *docorder_here << "\" to OID.\n\n";
|
|---|
| 192 |
|
|---|
| 193 | }
|
|---|
| 194 | else
|
|---|
| 195 | {
|
|---|
| 196 | docresultmap::iterator docset_here = queryresults.docs.docset.find (*docorder_here);
|
|---|
| 197 |
|
|---|
| 198 | // see if there is a result for this number,
|
|---|
| 199 | // if it is in the request set (or the request set is empty)
|
|---|
| 200 | if (docset_here != queryresults.docs.docset.end() && (request.docSet.empty() || in_set(request.docSet, trans_OID)))
|
|---|
| 201 | {
|
|---|
| 202 | // Now handled by Lucene directly
|
|---|
| 203 | //if (resultnum >= startresults) {
|
|---|
| 204 |
|
|---|
| 205 | // add this document
|
|---|
| 206 | resultdoc.OID = trans_OID;
|
|---|
| 207 | resultdoc.result_num = resultnum;
|
|---|
| 208 | resultdoc.ranking = (int)((*docset_here).second.docweight * 10000.0 + 0.5);
|
|---|
| 209 |
|
|---|
| 210 | response.docInfo.push_back (resultdoc);
|
|---|
| 211 | //}
|
|---|
| 212 | ++resultnum;
|
|---|
| 213 | }
|
|---|
| 214 | } // else
|
|---|
| 215 |
|
|---|
| 216 | ++docorder_here;
|
|---|
| 217 | }
|
|---|
| 218 | } // if need matching docs
|
|---|
| 219 |
|
|---|
| 220 | // assemble the term results
|
|---|
| 221 | if (need_term_info(request.filterResultOptions)) {
|
|---|
| 222 | // note: the terms have already been sorted and uniqued - ?? have they??
|
|---|
| 223 |
|
|---|
| 224 | TermInfo_t terminfo;
|
|---|
| 225 | bool terms_first = true;
|
|---|
| 226 |
|
|---|
| 227 | termfreqclassarray::iterator terms_here = queryresults.terms.begin();
|
|---|
| 228 | termfreqclassarray::iterator terms_end = queryresults.terms.end();
|
|---|
| 229 |
|
|---|
| 230 | while (terms_here != terms_end) {
|
|---|
| 231 | terminfo.clear();
|
|---|
| 232 | terminfo.term = (*terms_here).termstr;
|
|---|
| 233 | terminfo.freq = (*terms_here).termfreq;
|
|---|
| 234 | // lucene doesn't return any termvariants at this stage,
|
|---|
| 235 | // so make sure the original term is set
|
|---|
| 236 | terminfo.matchTerms.push_back(terminfo.term);
|
|---|
| 237 |
|
|---|
| 238 | // this bit gets the matchTerms ie the equivalent (stem/casefold) terms
|
|---|
| 239 | if (terms_first) {
|
|---|
| 240 | text_tset::iterator termvariants_here = queryresults.termvariants.begin();
|
|---|
| 241 | text_tset::iterator termvariants_end = queryresults.termvariants.end();
|
|---|
| 242 | while (termvariants_here != termvariants_end) {
|
|---|
| 243 | terminfo.matchTerms.push_back (*termvariants_here);
|
|---|
| 244 | ++termvariants_here;
|
|---|
| 245 | }
|
|---|
| 246 | }
|
|---|
| 247 | terms_first = false;
|
|---|
| 248 |
|
|---|
| 249 | response.termInfo.push_back (terminfo);
|
|---|
| 250 |
|
|---|
| 251 | ++terms_here;
|
|---|
| 252 | }
|
|---|
| 253 |
|
|---|
| 254 | // add the stop words
|
|---|
| 255 | text_tset::iterator stopwords_here = queryresults.stopwords.begin();
|
|---|
| 256 | text_tset::iterator stopwords_end = queryresults.stopwords.end();
|
|---|
| 257 | while (stopwords_here != stopwords_end) {
|
|---|
| 258 | response.stopwords.insert(*stopwords_here);
|
|---|
| 259 | ++stopwords_here;
|
|---|
| 260 | }
|
|---|
| 261 | }
|
|---|
| 262 |
|
|---|
| 263 | response.numDocs = queryresults.docs_matched;
|
|---|
| 264 | response.isApprox = queryresults.is_approx;
|
|---|
| 265 | }
|
|---|
| 266 |
|
|---|
| 267 | void lucenequeryfilterclass::browsefilter(const FilterRequest_t &request,
|
|---|
| 268 | FilterResponse_t &response,
|
|---|
| 269 | comerror_t &err, ostream &logout) {
|
|---|
| 270 |
|
|---|
| 271 | outconvertclass text_t2ascii;
|
|---|
| 272 |
|
|---|
| 273 | // get the query parameters
|
|---|
| 274 | int startresults, endresults;
|
|---|
| 275 | text_t phrasematch; // not used here any more, just have it so can use
|
|---|
| 276 | // parse_query_params function
|
|---|
| 277 |
|
|---|
| 278 | vector<queryparamclass> queryfilterparams;
|
|---|
| 279 | parse_query_params (request, queryfilterparams, startresults,
|
|---|
| 280 | endresults, phrasematch, logout);
|
|---|
| 281 |
|
|---|
| 282 | vector<queryparamclass>::const_iterator query_here = queryfilterparams.begin();
|
|---|
| 283 |
|
|---|
| 284 | // do query
|
|---|
| 285 | queryresultsclass queryresults;
|
|---|
| 286 | queryresults.clear();
|
|---|
| 287 |
|
|---|
| 288 | int numDocs = endresults-startresults;
|
|---|
| 289 | textsearchptr->setcollectdir (collectdir);
|
|---|
| 290 |
|
|---|
| 291 | if (!((lucenesearchclass*)textsearchptr)->browse_search((*query_here), startresults, numDocs, queryresults)) {
|
|---|
| 292 | // most likely a system problem
|
|---|
| 293 | logout << text_t2ascii
|
|---|
| 294 | << "system problem: could not do full text browse with lucene for index \""
|
|---|
| 295 | << (*query_here).index << (*query_here).subcollection
|
|---|
| 296 | << (*query_here).language << "\".\n\n";
|
|---|
| 297 | err = systemProblem;
|
|---|
| 298 | return;
|
|---|
| 299 | }
|
|---|
| 300 |
|
|---|
| 301 | // assemble the term results
|
|---|
| 302 | TermInfo_t terminfo;
|
|---|
| 303 |
|
|---|
| 304 | termfreqclassarray::iterator terms_here = queryresults.terms.begin();
|
|---|
| 305 | termfreqclassarray::iterator terms_end = queryresults.terms.end();
|
|---|
| 306 |
|
|---|
| 307 | while (terms_here != terms_end) {
|
|---|
| 308 | terminfo.clear();
|
|---|
| 309 | terminfo.term = (*terms_here).termstr;
|
|---|
| 310 | terminfo.freq = (*terms_here).termfreq;
|
|---|
| 311 |
|
|---|
| 312 | response.termInfo.push_back (terminfo);
|
|---|
| 313 |
|
|---|
| 314 | ++terms_here;
|
|---|
| 315 | }
|
|---|
| 316 |
|
|---|
| 317 |
|
|---|
| 318 | }
|
|---|
| 319 |
|
|---|
| 320 | // lucenesearchptr and gdbmptr are assumed to be valid
|
|---|
| 321 | void lucenequeryfilterclass::do_multi_query (const FilterRequest_t &request,
|
|---|
| 322 | const vector<queryparamclass> &query_params,
|
|---|
| 323 | queryresultsclass &multiresults,
|
|---|
| 324 | comerror_t &err, ostream &logout) {
|
|---|
| 325 | outconvertclass text_t2ascii;
|
|---|
| 326 |
|
|---|
| 327 | err = noError;
|
|---|
| 328 | textsearchptr->setcollectdir (collectdir);
|
|---|
| 329 | multiresults.clear();
|
|---|
| 330 |
|
|---|
| 331 | vector<queryparamclass>::const_iterator query_here = query_params.begin();
|
|---|
| 332 | vector<queryparamclass>::const_iterator query_end = query_params.end();
|
|---|
| 333 | while (query_here != query_end) {
|
|---|
| 334 | queryresultsclass thisqueryresults;
|
|---|
| 335 | if (!textsearchptr->search((*query_here), thisqueryresults)) {
|
|---|
| 336 | // most likely a system problem
|
|---|
| 337 | logout << text_t2ascii
|
|---|
| 338 | << "system problem: could not do search with lucene for index \""
|
|---|
| 339 | << (*query_here).index << (*query_here).level
|
|---|
| 340 | << (*query_here).subcollection
|
|---|
| 341 | << (*query_here).language << "\".\n\n";
|
|---|
| 342 | err = systemProblem;
|
|---|
| 343 | return;
|
|---|
| 344 | }
|
|---|
| 345 |
|
|---|
| 346 | // check for syntax error
|
|---|
| 347 | if (thisqueryresults.syntax_error==true) {
|
|---|
| 348 | logout << text_t2ascii
|
|---|
| 349 | << "syntax problem: invalid query string \""
|
|---|
| 350 | << (*query_here).querystring<<"\".\n";
|
|---|
| 351 | err = syntaxError;
|
|---|
| 352 | return;
|
|---|
| 353 | }
|
|---|
| 354 | // combine the results
|
|---|
| 355 | if (need_matching_docs (request.filterResultOptions)) {
|
|---|
| 356 |
|
|---|
| 357 | if (query_params.size() == 1) {
|
|---|
| 358 | multiresults.error_message = thisqueryresults.error_message;
|
|---|
| 359 | multiresults.docs = thisqueryresults.docs; // just one set of results
|
|---|
| 360 | multiresults.docs_matched = thisqueryresults.docs_matched;
|
|---|
| 361 | multiresults.is_approx = thisqueryresults.is_approx;
|
|---|
| 362 |
|
|---|
| 363 | } else {
|
|---|
| 364 | if ((*query_here).combinequery == "and") {
|
|---|
| 365 | multiresults.docs.combine_and (thisqueryresults.docs);
|
|---|
| 366 | } else if ((*query_here).combinequery == "or") {
|
|---|
| 367 | multiresults.docs.combine_or (thisqueryresults.docs);
|
|---|
| 368 | } else if ((*query_here).combinequery == "not") {
|
|---|
| 369 | multiresults.docs.combine_not (thisqueryresults.docs);
|
|---|
| 370 | }
|
|---|
| 371 | multiresults.docs_matched = multiresults.docs.docset.size();
|
|---|
| 372 | multiresults.is_approx = Exact;
|
|---|
| 373 | }
|
|---|
| 374 | }
|
|---|
| 375 |
|
|---|
| 376 | // combine the term information
|
|---|
| 377 | if (need_term_info (request.filterResultOptions)) {
|
|---|
| 378 | // append the terms
|
|---|
| 379 | multiresults.orgterms.insert(multiresults.orgterms.end(),
|
|---|
| 380 | thisqueryresults.orgterms.begin(),
|
|---|
| 381 | thisqueryresults.orgterms.end());
|
|---|
| 382 |
|
|---|
| 383 |
|
|---|
| 384 | // add the term variants -
|
|---|
| 385 | text_tset::iterator termvar_here = thisqueryresults.termvariants.begin();
|
|---|
| 386 | text_tset::iterator termvar_end = thisqueryresults.termvariants.end();
|
|---|
| 387 | while (termvar_here != termvar_end) {
|
|---|
| 388 | multiresults.termvariants.insert(*termvar_here);
|
|---|
| 389 | ++termvar_here;
|
|---|
| 390 | }
|
|---|
| 391 |
|
|---|
| 392 | // add the stop words
|
|---|
| 393 | text_tset::iterator stopwords_here = thisqueryresults.stopwords.begin();
|
|---|
| 394 | text_tset::iterator stopwords_end = thisqueryresults.stopwords.end();
|
|---|
| 395 | while (stopwords_here != stopwords_end) {
|
|---|
| 396 | multiresults.stopwords.insert(*stopwords_here);
|
|---|
| 397 | ++stopwords_here;
|
|---|
| 398 | }
|
|---|
| 399 | }
|
|---|
| 400 |
|
|---|
| 401 | ++query_here;
|
|---|
| 402 | }
|
|---|
| 403 |
|
|---|
| 404 | // sort and unique the query terms
|
|---|
| 405 | multiresults.sortuniqqueryterms ();
|
|---|
| 406 | }
|
|---|
| 407 |
|
|---|
| 408 |
|
|---|
| 409 |
|
|---|
