[1324] | 1 | /**********************************************************************
|
---|
| 2 | *
|
---|
| 3 | * mgppsearch.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 | #include "gsdlconf.h"
|
---|
| 28 | #include "mgppsearch.h"
|
---|
| 29 | #include "fileutil.h"
|
---|
| 30 | #include "GSDLQueryParser.h"
|
---|
| 31 | #include "MGQuery.h"
|
---|
| 32 | #include "TextGet.h"
|
---|
| 33 | #include "queryinfo.h"
|
---|
| 34 |
|
---|
| 35 |
|
---|
| 36 |
|
---|
| 37 | static text_t getindexsuffix(const queryparamclass &qp) {
|
---|
| 38 | text_t indexsuffix = "index";
|
---|
| 39 | text_t ind = qp.index;
|
---|
| 40 | text_t sub = qp.subcollection;
|
---|
| 41 | text_t lang = qp.language;
|
---|
| 42 |
|
---|
| 43 | indexsuffix = filename_cat(indexsuffix, ind + sub + lang, qp.collection);
|
---|
| 44 | return indexsuffix;
|
---|
| 45 |
|
---|
| 46 | }
|
---|
| 47 |
|
---|
| 48 | ////////////////////
|
---|
| 49 | // mgppsearch class //
|
---|
| 50 | ////////////////////
|
---|
| 51 |
|
---|
| 52 | mgppsearchclass::mgppsearchclass ()
|
---|
| 53 | : searchclass() {
|
---|
[1908] | 54 |
|
---|
| 55 | gdbm_level = "Document";
|
---|
[1324] | 56 | }
|
---|
| 57 |
|
---|
| 58 | mgppsearchclass::~mgppsearchclass ()
|
---|
| 59 | {
|
---|
| 60 | if (cache != NULL)
|
---|
| 61 | {
|
---|
| 62 | delete cache;
|
---|
| 63 | cache = NULL;
|
---|
| 64 | }
|
---|
[2697] | 65 |
|
---|
| 66 | if (indexData !=NULL) {
|
---|
| 67 | indexData->UnloadData();
|
---|
| 68 | delete indexData;
|
---|
| 69 | indexData = NULL;
|
---|
| 70 | }
|
---|
| 71 |
|
---|
[1324] | 72 | }
|
---|
| 73 |
|
---|
[1908] | 74 | void mgppsearchclass::set_gdbm_level(text_t &level) {
|
---|
| 75 | gdbm_level = level;
|
---|
| 76 |
|
---|
| 77 | }
|
---|
| 78 |
|
---|
[1324] | 79 | bool mgppsearchclass::search(const queryparamclass &queryparams,
|
---|
| 80 | queryresultsclass &queryresult) {
|
---|
[2701] | 81 |
|
---|
| 82 | #ifdef __WIN32__
|
---|
| 83 | char basepath[]="";
|
---|
| 84 | #else
|
---|
| 85 | char basepath[] = "/";
|
---|
| 86 | #endif
|
---|
| 87 |
|
---|
[2699] | 88 | char *indexname = (filename_cat(collectdir, getindexsuffix(queryparams))).getcstr();
|
---|
| 89 |
|
---|
| 90 | // load index data
|
---|
[2697] | 91 | if (indexData == NULL) {
|
---|
| 92 | indexData = new IndexData();
|
---|
| 93 | }
|
---|
| 94 | if (!indexData->LoadData (basepath, indexname)) {
|
---|
[1324] | 95 | cerr<<"couldn't load index data\n"<<endl;
|
---|
| 96 | return false;
|
---|
| 97 | }
|
---|
| 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 |
|
---|
[1771] | 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 |
|
---|
[1324] | 116 | // use default query info settings - change to reflect user preferences??
|
---|
| 117 | QueryInfo queryInfo;
|
---|
| 118 | SetCStr (queryInfo.docLevel, (queryparams.level.getcstr()));
|
---|
| 119 | queryInfo.maxDocs = (unsigned long)queryparams.maxdocs;
|
---|
[1771] | 120 | queryInfo.sortByRank = (queryparams.search_type == 1);
|
---|
[1324] | 121 | queryInfo.exactWeights = false;
|
---|
[1771] | 122 | queryInfo.needRankInfo = true; // used for overall term freq as well as ranking
|
---|
[1324] | 123 | queryInfo.needTermFreqs = true;
|
---|
| 124 |
|
---|
| 125 | ExtQueryResult queryResult;
|
---|
| 126 |
|
---|
| 127 | UCArray queryArray;
|
---|
| 128 | SetCStr(queryArray, (queryparams.querystring.getcstr()));
|
---|
| 129 |
|
---|
| 130 | // create the mgpp query tree
|
---|
| 131 | QueryNode *queryTree = NULL;
|
---|
[1771] | 132 | queryTree = ParseQuery(queryArray, defaultBoolCombine, defaultStemMethod);
|
---|
[1324] | 133 |
|
---|
| 134 | UCArray level;
|
---|
| 135 | UCArrayClear(level);
|
---|
| 136 |
|
---|
[1908] | 137 | //set the level for results
|
---|
| 138 | SetCStr(level, gdbm_level.getcstr());
|
---|
| 139 |
|
---|
[1324] | 140 |
|
---|
| 141 | // do the query
|
---|
[2697] | 142 | MGQuery(*indexData, queryInfo, queryTree, queryResult, level);
|
---|
[1324] | 143 |
|
---|
| 144 |
|
---|
| 145 | // convert ExtQueryResult to queryresultclass
|
---|
| 146 |
|
---|
| 147 | queryresult.docs_matched = (int)queryResult.docs.size();
|
---|
[1689] | 148 |
|
---|
| 149 | if (queryresult.docs_matched == (int)queryResult.actualNumDocs) {
|
---|
| 150 | queryresult.is_approx = Exact;
|
---|
| 151 | }
|
---|
| 152 | else if (queryresult.docs_matched < (int)queryResult.actualNumDocs) {
|
---|
| 153 | queryresult.is_approx = MoreThan;
|
---|
| 154 | }
|
---|
| 155 | else {
|
---|
| 156 | queryresult.is_approx = Approximate;
|
---|
| 157 | }
|
---|
| 158 |
|
---|
[1324] | 159 | docresultclass doc;
|
---|
| 160 | for (int i=0; i<(int)queryResult.docs.size(); i++) {
|
---|
| 161 | doc.clear();
|
---|
| 162 | doc.docnum = (int)queryResult.levels[i];
|
---|
| 163 | doc.docweight = queryResult.ranks[i];
|
---|
| 164 | queryresult.docs.docset[doc.docnum] = doc;
|
---|
| 165 | queryresult.docs.docorder.push_back(doc.docnum);
|
---|
| 166 |
|
---|
| 167 | }
|
---|
| 168 |
|
---|
| 169 | // term info
|
---|
| 170 | termfreqclass term;
|
---|
[2545] | 171 | for (int k=0; k<(int)queryResult.termFreqs.size(); k++) {
|
---|
[1324] | 172 | term.clear();
|
---|
[2545] | 173 | term.termstr = GetCStr(queryResult.termFreqs[k].term);
|
---|
[1324] | 174 | term.termstemstr = term.termstr;
|
---|
[2545] | 175 | term.termfreq = queryResult.termFreqs[k].termFreq;
|
---|
[1324] | 176 | queryresult.terms.push_back(term);
|
---|
[1834] | 177 | queryresult.orgterms.push_back(term); // should this change??
|
---|
| 178 |
|
---|
[2549] | 179 | for (int j=0; j<(int)queryResult.termFreqs[k].equivTerms.size(); j++) {
|
---|
| 180 | queryresult.termvariants.insert(GetCStr(queryResult.termFreqs[k].equivTerms[j]));
|
---|
[1834] | 181 | }
|
---|
[1324] | 182 |
|
---|
| 183 | }
|
---|
| 184 | // clean up
|
---|
| 185 | delete indexname;
|
---|
| 186 | return true;
|
---|
| 187 |
|
---|
| 188 | }
|
---|
| 189 |
|
---|
| 190 |
|
---|
| 191 | bool mgppsearchclass::browse_search(const queryparamclass &queryparams, int start, int numDocs,
|
---|
| 192 | queryresultsclass &queryresult) {
|
---|
| 193 |
|
---|
[2701] | 194 | #ifdef __WIN32__
|
---|
| 195 | char basepath[]="";
|
---|
| 196 | #else
|
---|
| 197 | char basepath[] = "/";
|
---|
| 198 | #endif
|
---|
| 199 |
|
---|
[2699] | 200 | char *indexname = (filename_cat(collectdir, getindexsuffix(queryparams))).getcstr();
|
---|
[1324] | 201 |
|
---|
[2697] | 202 | if (indexData == NULL) {
|
---|
| 203 | indexData = new IndexData();
|
---|
| 204 | }
|
---|
| 205 | if (!indexData->LoadData (basepath, indexname)) {
|
---|
[1324] | 206 | cerr<<"couldn't load index data\n"<<endl;
|
---|
| 207 | return false;
|
---|
| 208 | }
|
---|
| 209 |
|
---|
| 210 | UCArray level;
|
---|
| 211 | UCArrayClear(level);
|
---|
| 212 |
|
---|
| 213 | //browse always at top level
|
---|
| 214 | SetCStr(level, "Document");
|
---|
| 215 |
|
---|
| 216 |
|
---|
| 217 | BrowseQueryNode browseNode;
|
---|
| 218 | browseNode.startPosition = start;
|
---|
| 219 | browseNode.numTerms = numDocs;
|
---|
| 220 |
|
---|
| 221 | BrowseQueryResult browseResult;
|
---|
| 222 |
|
---|
| 223 | UCArrayClear(browseNode.term);
|
---|
| 224 | SetCStr(browseNode.term, (queryparams.querystring.getcstr()));
|
---|
| 225 |
|
---|
| 226 | // do the actual query
|
---|
[2697] | 227 | MGBrowseQuery(*indexData, level, browseNode, browseResult);
|
---|
[1324] | 228 |
|
---|
| 229 | // load results into term info
|
---|
| 230 | termfreqclass term;
|
---|
| 231 | for (int i=0; i<(int)browseResult.termFreqs.size(); i++) {
|
---|
| 232 | term.clear();
|
---|
| 233 | term.termstr = GetCStr(browseResult.termFreqs[i].term);
|
---|
| 234 | term.termstemstr = term.termstr;
|
---|
| 235 | term.termfreq = browseResult.termFreqs[i].termFreq;
|
---|
| 236 | queryresult.terms.push_back(term);
|
---|
| 237 | queryresult.orgterms.push_back(term);
|
---|
| 238 |
|
---|
| 239 | }
|
---|
| 240 | // clean up
|
---|
| 241 | delete indexname;
|
---|
| 242 |
|
---|
| 243 | return true;
|
---|
| 244 | }
|
---|
| 245 |
|
---|
| 246 | // the document text for 'docnum' is placed in 'output'
|
---|
| 247 | // docTargetDocument returns 'true' if it was able to
|
---|
| 248 | // try to get a document
|
---|
| 249 | // collection is needed to see if an index from the
|
---|
| 250 | // collection is loaded. THe default index bits are just there cos
|
---|
| 251 | // the mg version needs them
|
---|
| 252 |
|
---|
| 253 | bool mgppsearchclass::docTargetDocument(const text_t &/*defaultindex*/,
|
---|
| 254 | const text_t &/*defaultsubcollection*/,
|
---|
| 255 | const text_t &/*defaultlanguage*/,
|
---|
| 256 | const text_t &collection,
|
---|
| 257 | int docnum,
|
---|
| 258 | text_t &output) {
|
---|
| 259 |
|
---|
[2701] | 260 | #ifdef __WIN32__
|
---|
| 261 | char basepath[]="";
|
---|
| 262 | #else
|
---|
| 263 | char basepath[] = "/";
|
---|
| 264 | #endif
|
---|
| 265 | char *textname = (filename_cat(collectdir, "index", "text", collection)).getcstr();;
|
---|
[2699] | 266 |
|
---|
[1324] | 267 | TextData textdata;
|
---|
[2699] | 268 | if(!textdata.LoadData(basepath, textname)) {
|
---|
| 269 | cout<<"couldn't load text data\n"<<endl;
|
---|
[1324] | 270 | return false;
|
---|
| 271 | }
|
---|
| 272 | UCArray doctext;
|
---|
| 273 | UCArray level;
|
---|
[1908] | 274 | SetCStr(level, gdbm_level.getcstr());
|
---|
[1324] | 275 | if (!GetDocText(textdata, level, (unsigned long)docnum, doctext)) {
|
---|
[2699] | 276 | cout<<"couldn't retrieve document text\n";
|
---|
[1324] | 277 | return false;
|
---|
| 278 | }
|
---|
| 279 |
|
---|
| 280 | // convert UCArray to text_t
|
---|
| 281 | output.clear();
|
---|
| 282 | output = GetCStr(doctext);
|
---|
| 283 |
|
---|
| 284 | // here need to remove the <Document>, <Section>, <Paragraph> tags
|
---|
| 285 |
|
---|
| 286 | // mg converts to unicode, this may need to be added here???
|
---|
| 287 |
|
---|
| 288 | //clean up
|
---|
| 289 | textdata.UnloadData ();
|
---|
[2699] | 290 | delete textname;
|
---|
[1324] | 291 |
|
---|
| 292 | return true;
|
---|
| 293 |
|
---|
| 294 | }
|
---|
| 295 |
|
---|
[2679] | 296 | // used to clear any cached databases for persistent versions of
|
---|
| 297 | // Greenstone like the Windows local library
|
---|
| 298 | void mgppsearchclass::unload_database () {
|
---|
[1324] | 299 |
|
---|
[2699] | 300 | if (indexData !=NULL) {
|
---|
| 301 | indexData->UnloadData();
|
---|
| 302 | }
|
---|
[2679] | 303 | }
|
---|
[1324] | 304 |
|
---|
| 305 |
|
---|
| 306 |
|
---|
| 307 |
|
---|
[2679] | 308 |
|
---|