[270] | 1 | /**********************************************************************
|
---|
| 2 | *
|
---|
| 3 | * querytools.cpp --
|
---|
| 4 | * Copyright (C) 1999 The New Zealand Digital Library Project
|
---|
| 5 | *
|
---|
[533] | 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.
|
---|
[270] | 9 | *
|
---|
[533] | 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 | *
|
---|
[270] | 24 | *********************************************************************/
|
---|
| 25 |
|
---|
| 26 | #include "querytools.h"
|
---|
[1373] | 27 | #include <ctype.h>
|
---|
[270] | 28 |
|
---|
[759] | 29 | // request.filterResultOptions and request.fields (if required) should
|
---|
| 30 | // be set from the calling code
|
---|
| 31 | void set_queryfilter_options (FilterRequest_t &request, const text_t &querystring,
|
---|
| 32 | cgiargsclass &args) {
|
---|
[270] | 33 |
|
---|
| 34 | request.filterName = "QueryFilter";
|
---|
| 35 |
|
---|
| 36 | OptionValue_t option;
|
---|
[470] | 37 |
|
---|
[270] | 38 | option.name = "Term";
|
---|
[759] | 39 | option.value = querystring;
|
---|
[270] | 40 | request.filterOptions.push_back (option);
|
---|
| 41 |
|
---|
| 42 | option.name = "QueryType";
|
---|
| 43 | option.value = (args.getintarg("t")) ? "ranked" : "boolean";
|
---|
| 44 | request.filterOptions.push_back (option);
|
---|
| 45 |
|
---|
| 46 | option.name = "Casefold";
|
---|
| 47 | option.value = (args.getintarg("k")) ? "true" : "false";
|
---|
| 48 | request.filterOptions.push_back (option);
|
---|
| 49 |
|
---|
| 50 | option.name = "Stem";
|
---|
| 51 | option.value = (args.getintarg("s")) ? "true" : "false";
|
---|
| 52 | request.filterOptions.push_back (option);
|
---|
| 53 |
|
---|
| 54 | if (!args["h"].empty()) {
|
---|
| 55 | option.name = "Index";
|
---|
| 56 | option.value = args["h"];
|
---|
| 57 | request.filterOptions.push_back (option);
|
---|
| 58 | }
|
---|
| 59 |
|
---|
| 60 | if (!args["j"].empty()) {
|
---|
| 61 | option.name = "Subcollection";
|
---|
| 62 | option.value = args["j"];
|
---|
| 63 | request.filterOptions.push_back (option);
|
---|
| 64 | }
|
---|
| 65 |
|
---|
| 66 | if (!args["n"].empty()) {
|
---|
| 67 | option.name = "Language";
|
---|
| 68 | option.value = args["n"];
|
---|
| 69 | request.filterOptions.push_back (option);
|
---|
| 70 | }
|
---|
[1329] | 71 |
|
---|
| 72 | if (!args["g"].empty()) { // granularity for mgpp
|
---|
| 73 | option.name = "Level";
|
---|
| 74 | option.value = args["g"];
|
---|
| 75 | request.filterOptions.push_back (option);
|
---|
| 76 | }
|
---|
[270] | 77 |
|
---|
[759] | 78 | set_more_queryfilter_options (request, args);
|
---|
| 79 | }
|
---|
| 80 |
|
---|
| 81 | void set_queryfilter_options (FilterRequest_t &request, const text_t &querystring1,
|
---|
| 82 | const text_t &querystring2, cgiargsclass &args) {
|
---|
| 83 |
|
---|
| 84 | set_queryfilter_options (request, querystring1, args);
|
---|
| 85 |
|
---|
[349] | 86 | // fill in the second query if needed
|
---|
| 87 | if (!args["cq2"].empty()) {
|
---|
[759] | 88 | OptionValue_t option;
|
---|
| 89 |
|
---|
[349] | 90 | option.name = "CombineQuery";
|
---|
| 91 | option.value = args["cq2"];
|
---|
| 92 | request.filterOptions.push_back (option);
|
---|
| 93 |
|
---|
| 94 | option.name = "Term";
|
---|
[759] | 95 | option.value = querystring2;
|
---|
[349] | 96 | request.filterOptions.push_back (option);
|
---|
[759] | 97 |
|
---|
[349] | 98 | option.name = "QueryType";
|
---|
| 99 | option.value = (args.getintarg("t")) ? "ranked" : "boolean";
|
---|
| 100 | request.filterOptions.push_back (option);
|
---|
| 101 |
|
---|
| 102 | option.name = "Casefold";
|
---|
| 103 | option.value = (args.getintarg("k")) ? "true" : "false";
|
---|
| 104 | request.filterOptions.push_back (option);
|
---|
| 105 |
|
---|
| 106 | option.name = "Stem";
|
---|
| 107 | option.value = (args.getintarg("s")) ? "true" : "false";
|
---|
| 108 | request.filterOptions.push_back (option);
|
---|
| 109 |
|
---|
| 110 | if (!args["h2"].empty()) {
|
---|
| 111 | option.name = "Index";
|
---|
| 112 | option.value = args["h2"];
|
---|
| 113 | request.filterOptions.push_back (option);
|
---|
| 114 | }
|
---|
| 115 |
|
---|
| 116 | if (!args["j2"].empty()) {
|
---|
| 117 | option.name = "Subcollection";
|
---|
| 118 | option.value = args["j2"];
|
---|
| 119 | request.filterOptions.push_back (option);
|
---|
| 120 | }
|
---|
| 121 |
|
---|
| 122 | if (!args["n2"].empty()) {
|
---|
| 123 | option.name = "Language";
|
---|
| 124 | option.value = args["n2"];
|
---|
| 125 | request.filterOptions.push_back (option);
|
---|
| 126 | }
|
---|
| 127 | }
|
---|
[759] | 128 | set_more_queryfilter_options (request, args);
|
---|
| 129 | }
|
---|
[608] | 130 |
|
---|
[759] | 131 | void set_more_queryfilter_options (FilterRequest_t &request, cgiargsclass &args) {
|
---|
| 132 |
|
---|
| 133 | OptionValue_t option;
|
---|
[608] | 134 | int arg_m = args.getintarg("m");
|
---|
[759] | 135 |
|
---|
[608] | 136 | option.name = "Maxdocs";
|
---|
| 137 | option.value = arg_m;
|
---|
| 138 | request.filterOptions.push_back (option);
|
---|
[1329] | 139 |
|
---|
[759] | 140 | // option.name = "StartResults";
|
---|
| 141 | // option.value = args["r"];
|
---|
| 142 | // request.filterOptions.push_back (option);
|
---|
[270] | 143 |
|
---|
[759] | 144 | // option.name = "EndResults";
|
---|
| 145 | // int endresults = args.getintarg("o") + (args.getintarg("r") - 1);
|
---|
| 146 | // if ((endresults > arg_m) && (arg_m != -1)) endresults = arg_m;
|
---|
| 147 | // option.value = endresults;
|
---|
| 148 | // request.filterOptions.push_back (option);
|
---|
[270] | 149 | }
|
---|
| 150 |
|
---|
[470] | 151 | void format_querystring (text_t &querystring, int querymode) {
|
---|
[270] | 152 | text_t formattedstring;
|
---|
| 153 |
|
---|
| 154 | text_t::const_iterator here = querystring.begin();
|
---|
| 155 | text_t::const_iterator end = querystring.end();
|
---|
| 156 |
|
---|
| 157 | // space is used to insert spaces between Chinese
|
---|
| 158 | // characters. No space is needed before the first
|
---|
| 159 | // Chinese character.
|
---|
| 160 | bool space = false;
|
---|
| 161 |
|
---|
| 162 | // want to remove ()|!& from querystring so boolean queries are just
|
---|
[470] | 163 | // "all the words" queries (unless querymode is advanced)
|
---|
[270] | 164 | while (here != end) {
|
---|
[470] | 165 | if ((querymode == 0) && (*here == '(' || *here == ')' || *here == '|' ||
|
---|
| 166 | *here == '!' || *here == '&')) {
|
---|
[270] | 167 | formattedstring.push_back(' ');
|
---|
| 168 | } else {
|
---|
[397] | 169 | if ((*here >= 0x4e00 && *here <= 0x9fa5) ||
|
---|
| 170 | (*here >= 0xf900 && *here <= 0xfa2d)) {
|
---|
| 171 | // Chinese character
|
---|
| 172 | if (space) formattedstring.push_back (0x200b);
|
---|
| 173 | formattedstring.push_back (*here);
|
---|
| 174 | formattedstring.push_back (0x200b);
|
---|
| 175 | space = true;
|
---|
[270] | 176 | } else {
|
---|
[397] | 177 | // non-Chinese character
|
---|
| 178 | formattedstring.push_back (*here);
|
---|
| 179 | space = false;
|
---|
[270] | 180 | }
|
---|
| 181 | }
|
---|
| 182 | here ++;
|
---|
| 183 | }
|
---|
[397] | 184 | querystring = formattedstring;
|
---|
[270] | 185 | }
|
---|
| 186 |
|
---|
[1373] | 187 | void add_ands(text_t& querystring, int querytype)
|
---|
| 188 | {
|
---|
| 189 |
|
---|
| 190 | if(querytype==0)
|
---|
| 191 | {
|
---|
| 192 | text_t formattedstring;
|
---|
| 193 |
|
---|
| 194 | bool spacelast = false;
|
---|
| 195 |
|
---|
| 196 | text_t::const_iterator here = querystring.begin();
|
---|
| 197 | text_t::const_iterator end = querystring.end();
|
---|
| 198 |
|
---|
| 199 |
|
---|
| 200 | querystring = formattedstring;
|
---|
| 201 |
|
---|
| 202 | while(here!=end && isspace((*here)))
|
---|
| 203 | here++;
|
---|
| 204 |
|
---|
| 205 | if(here==end)
|
---|
| 206 | return;
|
---|
| 207 |
|
---|
| 208 | while(here!=end)
|
---|
| 209 | {
|
---|
| 210 | if(isspace((*here)))
|
---|
| 211 | {
|
---|
| 212 | while(here!=end && isspace((*here)))
|
---|
| 213 | here++;
|
---|
| 214 | if(here!=end)
|
---|
| 215 | formattedstring.appendcstr(" AND ");
|
---|
| 216 | }
|
---|
| 217 | if(!isspace((*here)))
|
---|
| 218 | {
|
---|
| 219 | formattedstring.push_back((*here));
|
---|
| 220 | }
|
---|
| 221 | here++;
|
---|
| 222 | }
|
---|
| 223 | querystring = formattedstring;
|
---|
| 224 | }
|
---|
| 225 | }
|
---|
| 226 |
|
---|
| 227 |
|
---|
| 228 |
|
---|
| 229 | void add_dates(text_t &querystring, int startdate, int enddate,
|
---|
| 230 | int startbc, int endbc)
|
---|
| 231 | {
|
---|
| 232 | if(startdate)
|
---|
| 233 | {
|
---|
| 234 | int querystringis = 0;
|
---|
| 235 | text_t::const_iterator here = querystring.begin();
|
---|
| 236 | text_t::const_iterator end = querystring.end();
|
---|
| 237 | while(here!=end)
|
---|
| 238 | {
|
---|
| 239 | if(!(isspace((*here)))){
|
---|
| 240 | here = end;
|
---|
| 241 | querystringis = 1;
|
---|
| 242 | }
|
---|
| 243 | else
|
---|
| 244 | here++;
|
---|
| 245 | }
|
---|
| 246 | //converting BCE dates
|
---|
| 247 | if(startbc && startdate > 0)
|
---|
| 248 | {
|
---|
| 249 | startdate *= -1;
|
---|
| 250 | }
|
---|
| 251 | if(endbc && enddate > 0)
|
---|
| 252 | {
|
---|
| 253 | enddate *= -1;
|
---|
| 254 | }
|
---|
| 255 | if(enddate != 0 && enddate<startdate)
|
---|
| 256 | {
|
---|
| 257 | cout<<"enddate too small"<<endl;
|
---|
| 258 | return;
|
---|
| 259 | }
|
---|
| 260 | if(querystringis)
|
---|
| 261 | querystring.appendcstr(" AND");
|
---|
| 262 | if(!enddate)
|
---|
| 263 | {
|
---|
| 264 | querystring.appendcstr(" [");
|
---|
| 265 | querystring.appendint(startdate);
|
---|
| 266 | querystring.appendcstr("]:Date");
|
---|
| 267 | }
|
---|
| 268 | else{
|
---|
| 269 | int nextdate = startdate;
|
---|
| 270 | querystring.appendcstr(" (");
|
---|
| 271 | while(nextdate<=enddate)
|
---|
| 272 | {
|
---|
| 273 | if(nextdate!=0)
|
---|
| 274 | {
|
---|
| 275 | querystring.appendcstr(" [");
|
---|
| 276 | querystring.appendint(nextdate);
|
---|
| 277 | querystring.appendcstr("]:Date");
|
---|
| 278 | }
|
---|
| 279 | nextdate++;
|
---|
| 280 | }
|
---|
| 281 | querystring.appendcstr(" )");
|
---|
| 282 | }
|
---|
| 283 | }
|
---|
| 284 | }
|
---|
[403] | 285 | void get_phrases (const text_t &querystring, text_tarray &phrases) {
|
---|
| 286 |
|
---|
| 287 | phrases.erase (phrases.begin(), phrases.end());
|
---|
| 288 | if (!querystring.empty()) {
|
---|
| 289 |
|
---|
| 290 | text_t::const_iterator end = querystring.end();
|
---|
| 291 | text_t::const_iterator here = findchar (querystring.begin(), end, '"');
|
---|
| 292 | if (here != end) {
|
---|
| 293 | text_t tmptext;
|
---|
| 294 | bool foundquote = false;
|
---|
| 295 | while (here != end) {
|
---|
| 296 | if (*here == '"') {
|
---|
| 297 | if (foundquote) {
|
---|
| 298 | if (!tmptext.empty()) {
|
---|
| 299 | phrases.push_back(tmptext);
|
---|
| 300 | tmptext.clear();
|
---|
| 301 | }
|
---|
| 302 | foundquote = false;
|
---|
| 303 | } else foundquote = true;
|
---|
| 304 | } else {
|
---|
| 305 | if (foundquote) tmptext.push_back (*here);
|
---|
| 306 | }
|
---|
| 307 | here ++;
|
---|
| 308 | }
|
---|
| 309 | }
|
---|
| 310 | }
|
---|
| 311 | }
|
---|