[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>
|
---|
[1914] | 28 | #include "unitool.h" // for is_unicode_letdig
|
---|
[270] | 29 |
|
---|
[12784] | 30 | // sets the ct, qt, qto arguments
|
---|
[11987] | 31 | void set_query_type_args(ColInfoResponse_t *cinfo, cgiargsclass &args) {
|
---|
| 32 |
|
---|
| 33 | if (args["ct"].empty()) {
|
---|
| 34 | text_t build_type = cinfo->buildType;
|
---|
| 35 | if (build_type == "mgpp") {
|
---|
| 36 | args["ct"] = "1";
|
---|
| 37 | } else if (build_type == "lucene") {
|
---|
| 38 | args["ct"] = "2";
|
---|
| 39 | } else {
|
---|
| 40 | args["ct"] = "0";
|
---|
| 41 | }
|
---|
| 42 | }
|
---|
| 43 | text_t arg_ct = args["ct"];
|
---|
| 44 | if (arg_ct == "0") {
|
---|
| 45 | // mg
|
---|
| 46 | args["qt"] = "0";
|
---|
| 47 | args["qto"] = "0";
|
---|
| 48 | return;
|
---|
| 49 | }
|
---|
| 50 |
|
---|
| 51 | if (!args["qt"].empty() && !args["qto"].empty()) {
|
---|
| 52 | return;
|
---|
| 53 | }
|
---|
| 54 |
|
---|
| 55 | text_tmap::iterator check = cinfo->format.find("SearchTypes");
|
---|
[12784] | 56 | text_t search_types;
|
---|
| 57 | if(check != cinfo->format.end() && !(*check).second.empty()){
|
---|
[11987] | 58 | search_types = (*check).second;
|
---|
[12784] | 59 | } else {
|
---|
| 60 | // assume plain,form
|
---|
| 61 | if (args["qto"].empty()) args["qto"] = "3";
|
---|
| 62 | if (args["qt"].empty()) {
|
---|
| 63 | int arg_qto = args.getintarg("qto");
|
---|
[12930] | 64 | if (arg_qto == 2) {
|
---|
[12784] | 65 | args["qt"] = "1";
|
---|
| 66 | } else {
|
---|
| 67 | args["qt"] = "0";
|
---|
| 68 | }
|
---|
[11987] | 69 | }
|
---|
[12784] | 70 | return;
|
---|
[11987] | 71 | }
|
---|
| 72 |
|
---|
[12784] | 73 |
|
---|
[11987] | 74 | if (args["qto"].empty()) {
|
---|
| 75 | unsigned int type = 0;
|
---|
| 76 | if (findword(search_types.begin(), search_types.end(), "form") != search_types.end()) {
|
---|
| 77 | type |= 2;
|
---|
| 78 | }
|
---|
| 79 | if (findword(search_types.begin(), search_types.end(), "plain") != search_types.end()) {
|
---|
| 80 | type |= 1;
|
---|
| 81 | }
|
---|
| 82 | args.setintarg("qto", type);
|
---|
| 83 | }
|
---|
| 84 |
|
---|
| 85 | if (args["qt"].empty()) {
|
---|
| 86 | int arg_qto = args.getintarg("qto");
|
---|
| 87 | if (arg_qto == 2 || (arg_qto == 3 && starts_with(search_types, "form"))) {
|
---|
| 88 | args["qt"] = "1";
|
---|
| 89 | } else {
|
---|
| 90 | args["qt"] = "0";
|
---|
| 91 | }
|
---|
| 92 | }
|
---|
| 93 | }
|
---|
| 94 |
|
---|
[12864] | 95 | // sets the ks, ss, afs (casesupport, stemsupport, accentfoldsupport) args
|
---|
| 96 | void set_stem_index_args(ColInfoResponse_t *cinfo, cgiargsclass &args) {
|
---|
| 97 | int stemIndexes = cinfo->stemIndexes;
|
---|
| 98 |
|
---|
| 99 | if (stemIndexes & SIcasefold) {
|
---|
| 100 | args["ks"] = 1;
|
---|
| 101 | }
|
---|
| 102 | if (stemIndexes & SIstem) {
|
---|
| 103 | args["ss"] = 1;
|
---|
| 104 | }
|
---|
| 105 | if (stemIndexes & SIaccentfold) {
|
---|
| 106 | args["afs"] = 1;
|
---|
| 107 | }
|
---|
| 108 |
|
---|
| 109 | }
|
---|
| 110 |
|
---|
[759] | 111 | // request.filterResultOptions and request.fields (if required) should
|
---|
| 112 | // be set from the calling code
|
---|
[12784] | 113 | void set_queryfilter_options (FilterRequest_t &request,
|
---|
| 114 | const text_t &querystring,
|
---|
[759] | 115 | cgiargsclass &args) {
|
---|
[270] | 116 |
|
---|
| 117 | request.filterName = "QueryFilter";
|
---|
| 118 |
|
---|
| 119 | OptionValue_t option;
|
---|
[470] | 120 |
|
---|
[270] | 121 | option.name = "Term";
|
---|
[759] | 122 | option.value = querystring;
|
---|
[270] | 123 | request.filterOptions.push_back (option);
|
---|
| 124 |
|
---|
| 125 | option.name = "QueryType";
|
---|
| 126 | option.value = (args.getintarg("t")) ? "ranked" : "boolean";
|
---|
| 127 | request.filterOptions.push_back (option);
|
---|
| 128 |
|
---|
[1774] | 129 | option.name = "MatchMode";
|
---|
[11765] | 130 | // mgpp in advanced mode, always use some query
|
---|
[12428] | 131 | if (args.getintarg("ct") == 1 && args.getintarg("b") == 1) {
|
---|
[11765] | 132 | option.value = "some";
|
---|
| 133 | } else {
|
---|
| 134 | option.value = (args.getintarg("t")) ? "some" : "all";
|
---|
| 135 | }
|
---|
[1774] | 136 | request.filterOptions.push_back (option);
|
---|
| 137 |
|
---|
[270] | 138 | option.name = "Casefold";
|
---|
| 139 | option.value = (args.getintarg("k")) ? "true" : "false";
|
---|
| 140 | request.filterOptions.push_back (option);
|
---|
| 141 |
|
---|
| 142 | option.name = "Stem";
|
---|
| 143 | option.value = (args.getintarg("s")) ? "true" : "false";
|
---|
| 144 | request.filterOptions.push_back (option);
|
---|
| 145 |
|
---|
[12864] | 146 | option.name = "AccentFold";
|
---|
| 147 | option.value = (args.getintarg("af")) ? "true" : "false";
|
---|
| 148 | request.filterOptions.push_back (option);
|
---|
| 149 |
|
---|
[270] | 150 | if (!args["h"].empty()) {
|
---|
| 151 | option.name = "Index";
|
---|
| 152 | option.value = args["h"];
|
---|
| 153 | request.filterOptions.push_back (option);
|
---|
| 154 | }
|
---|
| 155 |
|
---|
| 156 | if (!args["j"].empty()) {
|
---|
| 157 | option.name = "Subcollection";
|
---|
| 158 | option.value = args["j"];
|
---|
| 159 | request.filterOptions.push_back (option);
|
---|
| 160 | }
|
---|
| 161 |
|
---|
| 162 | if (!args["n"].empty()) {
|
---|
| 163 | option.name = "Language";
|
---|
| 164 | option.value = args["n"];
|
---|
| 165 | request.filterOptions.push_back (option);
|
---|
| 166 | }
|
---|
[1329] | 167 |
|
---|
| 168 | if (!args["g"].empty()) { // granularity for mgpp
|
---|
| 169 | option.name = "Level";
|
---|
| 170 | option.value = args["g"];
|
---|
| 171 | request.filterOptions.push_back (option);
|
---|
| 172 | }
|
---|
[270] | 173 |
|
---|
[12410] | 174 | if (!args["fs"].empty()) { // filter string for lucene
|
---|
| 175 | option.name = "FilterString";
|
---|
| 176 | option.value = args["fs"];
|
---|
| 177 | request.filterOptions.push_back (option);
|
---|
| 178 | }
|
---|
| 179 |
|
---|
[12276] | 180 | if (!args["sf"].empty()) { // sort field for lucene
|
---|
| 181 | option.name = "SortField";
|
---|
| 182 | option.value = args["sf"];
|
---|
| 183 | request.filterOptions.push_back (option);
|
---|
| 184 | }
|
---|
| 185 |
|
---|
[12771] | 186 | if (!args["fuzziness"].empty() && args["fuzziness"] != "100") { // fuzziness value for lucene
|
---|
[12770] | 187 | option.name = "Fuzziness";
|
---|
[12771] | 188 | option.value = (text_t) "0." + args["fuzziness"];
|
---|
[12770] | 189 | request.filterOptions.push_back (option);
|
---|
| 190 | }
|
---|
[12388] | 191 |
|
---|
[759] | 192 | set_more_queryfilter_options (request, args);
|
---|
| 193 | }
|
---|
| 194 |
|
---|
[12784] | 195 | void set_queryfilter_options (FilterRequest_t &request,
|
---|
| 196 | const text_t &querystring1,
|
---|
[759] | 197 | const text_t &querystring2, cgiargsclass &args) {
|
---|
| 198 |
|
---|
| 199 | set_queryfilter_options (request, querystring1, args);
|
---|
| 200 |
|
---|
[349] | 201 | // fill in the second query if needed
|
---|
| 202 | if (!args["cq2"].empty()) {
|
---|
[759] | 203 | OptionValue_t option;
|
---|
| 204 |
|
---|
[349] | 205 | option.name = "CombineQuery";
|
---|
| 206 | option.value = args["cq2"];
|
---|
| 207 | request.filterOptions.push_back (option);
|
---|
| 208 |
|
---|
| 209 | option.name = "Term";
|
---|
[759] | 210 | option.value = querystring2;
|
---|
[349] | 211 | request.filterOptions.push_back (option);
|
---|
[759] | 212 |
|
---|
[349] | 213 | option.name = "QueryType";
|
---|
| 214 | option.value = (args.getintarg("t")) ? "ranked" : "boolean";
|
---|
| 215 | request.filterOptions.push_back (option);
|
---|
| 216 |
|
---|
| 217 | option.name = "Casefold";
|
---|
| 218 | option.value = (args.getintarg("k")) ? "true" : "false";
|
---|
| 219 | request.filterOptions.push_back (option);
|
---|
| 220 |
|
---|
| 221 | option.name = "Stem";
|
---|
| 222 | option.value = (args.getintarg("s")) ? "true" : "false";
|
---|
| 223 | request.filterOptions.push_back (option);
|
---|
| 224 |
|
---|
[12864] | 225 | option.name = "AccentFold";
|
---|
| 226 | option.value = (args.getintarg("af")) ? "true" : "false";
|
---|
| 227 | request.filterOptions.push_back (option);
|
---|
| 228 |
|
---|
[349] | 229 | if (!args["h2"].empty()) {
|
---|
| 230 | option.name = "Index";
|
---|
| 231 | option.value = args["h2"];
|
---|
| 232 | request.filterOptions.push_back (option);
|
---|
| 233 | }
|
---|
| 234 |
|
---|
| 235 | if (!args["j2"].empty()) {
|
---|
| 236 | option.name = "Subcollection";
|
---|
| 237 | option.value = args["j2"];
|
---|
| 238 | request.filterOptions.push_back (option);
|
---|
| 239 | }
|
---|
| 240 |
|
---|
| 241 | if (!args["n2"].empty()) {
|
---|
| 242 | option.name = "Language";
|
---|
| 243 | option.value = args["n2"];
|
---|
| 244 | request.filterOptions.push_back (option);
|
---|
| 245 | }
|
---|
| 246 | }
|
---|
[759] | 247 | set_more_queryfilter_options (request, args);
|
---|
| 248 | }
|
---|
[608] | 249 |
|
---|
[12784] | 250 | void set_more_queryfilter_options (FilterRequest_t &request,
|
---|
| 251 | cgiargsclass &args) {
|
---|
[759] | 252 |
|
---|
| 253 | OptionValue_t option;
|
---|
[608] | 254 | int arg_m = args.getintarg("m");
|
---|
[759] | 255 |
|
---|
[608] | 256 | option.name = "Maxdocs";
|
---|
| 257 | option.value = arg_m;
|
---|
| 258 | request.filterOptions.push_back (option);
|
---|
[1329] | 259 |
|
---|
[759] | 260 | // option.name = "StartResults";
|
---|
| 261 | // option.value = args["r"];
|
---|
| 262 | // request.filterOptions.push_back (option);
|
---|
[270] | 263 |
|
---|
[759] | 264 | // option.name = "EndResults";
|
---|
| 265 | // int endresults = args.getintarg("o") + (args.getintarg("r") - 1);
|
---|
| 266 | // if ((endresults > arg_m) && (arg_m != -1)) endresults = arg_m;
|
---|
| 267 | // option.value = endresults;
|
---|
| 268 | // request.filterOptions.push_back (option);
|
---|
[270] | 269 | }
|
---|
| 270 |
|
---|
[11987] | 271 | bool is_special_character(int indexer_type, unsigned short character) {
|
---|
| 272 | // mgpp
|
---|
| 273 | if (indexer_type == 1) {
|
---|
| 274 | return (character == '#' || character == '/' || character == '*');
|
---|
| 275 | }
|
---|
| 276 | // lucene
|
---|
[12784] | 277 | else if (indexer_type == 2) {
|
---|
[11987] | 278 | return (character == '?' || character == '*' || character == '~' ||
|
---|
| 279 | character == '^');
|
---|
| 280 | }
|
---|
| 281 | return false;
|
---|
| 282 | }
|
---|
| 283 |
|
---|
[12784] | 284 | // This function removes boolean operators from simple searches, and segments
|
---|
| 285 | // chinese characters if segment=true
|
---|
[6584] | 286 | void format_querystring (text_t &querystring, int querymode, bool segment) {
|
---|
[270] | 287 | text_t formattedstring;
|
---|
| 288 |
|
---|
[12784] | 289 | // advanced search, no segmenting, don't need to do anything
|
---|
[6584] | 290 | if (querymode == 1 && !segment) return;
|
---|
| 291 |
|
---|
[270] | 292 | text_t::const_iterator here = querystring.begin();
|
---|
| 293 | text_t::const_iterator end = querystring.end();
|
---|
| 294 |
|
---|
| 295 | // space is used to insert spaces between Chinese
|
---|
| 296 | // characters. No space is needed before the first
|
---|
| 297 | // Chinese character.
|
---|
| 298 | bool space = false;
|
---|
| 299 |
|
---|
| 300 | // want to remove ()|!& from querystring so boolean queries are just
|
---|
[470] | 301 | // "all the words" queries (unless querymode is advanced)
|
---|
[270] | 302 | while (here != end) {
|
---|
[470] | 303 | if ((querymode == 0) && (*here == '(' || *here == ')' || *here == '|' ||
|
---|
| 304 | *here == '!' || *here == '&')) {
|
---|
[270] | 305 | formattedstring.push_back(' ');
|
---|
[6584] | 306 | } else if (segment) {
|
---|
[16980] | 307 | if ((*here >= 0x2e80 && *here <= 0xd7a3) ||
|
---|
| 308 | ( *here >= 0xf900 && *here <= 0xfa6a)) {
|
---|
| 309 | /* text_t not big enough to handle these. */
|
---|
| 310 | /* (*here >= 0x20000 && *here <= 0x2a6d6) ||
|
---|
| 311 | (*here >= 0x2f800 && *here <= 0x2fa1d)) { */
|
---|
[16645] | 312 |
|
---|
| 313 | // CJK character
|
---|
[8715] | 314 | if (!space) formattedstring.push_back (0x200b); // zero width space
|
---|
[397] | 315 | formattedstring.push_back (*here);
|
---|
| 316 | formattedstring.push_back (0x200b);
|
---|
| 317 | space = true;
|
---|
[270] | 318 | } else {
|
---|
[8715] | 319 |
|
---|
[397] | 320 | // non-Chinese character
|
---|
| 321 | formattedstring.push_back (*here);
|
---|
| 322 | space = false;
|
---|
[8715] | 323 |
|
---|
[270] | 324 | }
|
---|
[6584] | 325 |
|
---|
| 326 | } else {
|
---|
| 327 | formattedstring.push_back (*here);
|
---|
[270] | 328 | }
|
---|
[9620] | 329 | ++here;
|
---|
[270] | 330 | }
|
---|
[397] | 331 | querystring = formattedstring;
|
---|
[270] | 332 | }
|
---|
| 333 |
|
---|
[1373] | 334 |
|
---|
| 335 |
|
---|
[1467] | 336 |
|
---|
[3160] | 337 | // search history tool
|
---|
| 338 | // also used for form query macros
|
---|
[1914] | 339 | text_t escape_quotes(const text_t &querystring) {
|
---|
| 340 |
|
---|
| 341 | text_t::const_iterator here = querystring.begin();
|
---|
| 342 | text_t::const_iterator end = querystring.end();
|
---|
| 343 |
|
---|
| 344 | text_t escquery = "";
|
---|
| 345 | while (here != end) {
|
---|
[1988] | 346 | if (*here != '\'' && *here != '\"' && *here != '\n' && *here != '\r') escquery.push_back(*here);
|
---|
| 347 | else if (*here == '\n' || *here == '\r') {
|
---|
| 348 | escquery.push_back(' ');
|
---|
| 349 | } else {
|
---|
[1914] | 350 | escquery +="\\\\";
|
---|
| 351 | escquery.push_back(*here);
|
---|
| 352 | }
|
---|
| 353 |
|
---|
[9620] | 354 | ++here;
|
---|
[1914] | 355 | }
|
---|
| 356 | return escquery;
|
---|
| 357 |
|
---|
| 358 | }
|
---|
| 359 |
|
---|
[12784] | 360 | // Parses the terms into words, and adds #si if necessary
|
---|
| 361 | text_t addstemcase(const text_t &terms, const text_t &stem, const text_t &fold,
|
---|
| 362 | const int indexer_type) {
|
---|
| 363 |
|
---|
| 364 | // the default stem and case are set to 0 if this is being used, so we are only adding on qualifiers if stem or fold is 1.
|
---|
| 365 | if (stem == "0" && fold == "0") {
|
---|
[12791] | 366 | return terms;
|
---|
[12784] | 367 | }
|
---|
| 368 | // this is only for mgpp collections, shouldn't be called for anything else
|
---|
| 369 | if (indexer_type != 1) {
|
---|
[12791] | 370 | return terms;
|
---|
[12784] | 371 | }
|
---|
| 372 |
|
---|
| 373 | text_t outtext;
|
---|
| 374 | text_t word;
|
---|
| 375 |
|
---|
| 376 | text_t::const_iterator here = terms.begin();
|
---|
| 377 | text_t::const_iterator end = terms.end();
|
---|
| 378 |
|
---|
| 379 | while (here !=end) {
|
---|
| 380 |
|
---|
| 381 | if (is_unicode_letdig(*here) || is_special_character(indexer_type, *here)) {
|
---|
| 382 | // not word boundary
|
---|
| 383 | word.push_back(*here);
|
---|
| 384 | ++here;
|
---|
| 385 | }
|
---|
| 386 | else {
|
---|
| 387 | // found word boundary
|
---|
| 388 | if (!word.empty() ) {
|
---|
| 389 | if (starts_with(word, "NEAR") || starts_with(word, "WITHIN")) {
|
---|
| 390 | outtext += word;
|
---|
| 391 | word.clear();
|
---|
| 392 | }
|
---|
| 393 | else {
|
---|
| 394 | word += "#";
|
---|
| 395 | if (stem == "1") word += "s";
|
---|
| 396 | if (fold == "1") word += "i";
|
---|
| 397 | outtext += word;
|
---|
| 398 | word.clear();
|
---|
| 399 | }
|
---|
| 400 | }
|
---|
| 401 | // this only used in advanced form, so we leave in boolean operators
|
---|
[12792] | 402 | if (*here == '\"' || *here == '&' || *here == '|' || *here == '!' ||
|
---|
| 403 | *here == '(' || *here == ')' || is_unicode_space(*here)) {
|
---|
[12784] | 404 | outtext.push_back(*here);
|
---|
| 405 | }
|
---|
| 406 | ++here;
|
---|
| 407 | }
|
---|
| 408 | }
|
---|
| 409 |
|
---|
| 410 | // get last word
|
---|
| 411 | if (!word.empty()) {
|
---|
| 412 | word += "#";
|
---|
| 413 | if (stem == "1") word += "s";
|
---|
| 414 | if (fold == "1") word += "i";
|
---|
| 415 | word += " ";
|
---|
| 416 | outtext += word;
|
---|
| 417 | }
|
---|
| 418 | return outtext;
|
---|
| 419 | }
|
---|
| 420 |
|
---|
| 421 |
|
---|
[11765] | 422 | // some query form parsing functions for use with mgpp & lucene
|
---|
[1914] | 423 |
|
---|
[12784] | 424 | void parse_reg_query_form(text_t &querystring, cgiargsclass &args, bool segment)
|
---|
[8029] | 425 | {
|
---|
| 426 | querystring.clear();
|
---|
[1914] | 427 |
|
---|
[12784] | 428 | int argct = args.getintarg("ct");
|
---|
[8029] | 429 | int argt = args.getintarg("t");// t=0 -and, t=1 - or
|
---|
[12784] | 430 | int argb = args.getintarg("b");
|
---|
| 431 |
|
---|
| 432 | text_t combine;
|
---|
[8029] | 433 |
|
---|
[12784] | 434 | // lucene uses global combine, so only need this for mgpp
|
---|
| 435 | if (argct==1) {
|
---|
[8029] | 436 | if (argt == 0) combine = "&";
|
---|
| 437 | else combine = "|";
|
---|
| 438 | }
|
---|
[1914] | 439 |
|
---|
| 440 | text_t field = args["fqf"];
|
---|
| 441 | if (field.empty()) return; // no query
|
---|
| 442 | text_tarray fields;
|
---|
| 443 | splitchar(field.begin(), field.end(), ',', fields);
|
---|
| 444 |
|
---|
| 445 | text_t value = args["fqv"];
|
---|
| 446 | if (value.empty()) return; // somethings wrong
|
---|
| 447 | text_tarray values;
|
---|
| 448 | splitchar(value.begin(), value.end(), ',', values);
|
---|
| 449 |
|
---|
[8029] | 450 |
|
---|
[9620] | 451 | for (int i=0; i< values.size(); ++i) {
|
---|
[1914] | 452 | if (!values[i].empty()) {
|
---|
[12784] | 453 | text_t this_value = values[i];
|
---|
| 454 | // remove operators for simple search, segments text if necessary
|
---|
| 455 | format_querystring(this_value, argb, segment);
|
---|
| 456 | // add tag info for this field (and other processing)
|
---|
| 457 | format_field_info(this_value, fields[i], argct, argt, argb);
|
---|
| 458 | // add into query string
|
---|
| 459 | if (argct == 2) {
|
---|
| 460 | // lucene
|
---|
| 461 | // we don't worry about AND/OR, cos this is done by defaultcombineoperator
|
---|
| 462 | querystring += this_value+" ";
|
---|
| 463 | } else {
|
---|
| 464 | // mgpp
|
---|
| 465 | if (!querystring.empty()) {
|
---|
| 466 | querystring += " "+ combine+ " ";
|
---|
| 467 | }
|
---|
| 468 | querystring += this_value;
|
---|
[8029] | 469 | }
|
---|
[1914] | 470 | }
|
---|
| 471 | }
|
---|
| 472 | }
|
---|
| 473 |
|
---|
| 474 |
|
---|
[12784] | 475 | void parse_adv_query_form(text_t &querystring, cgiargsclass &args, bool segment){
|
---|
[1914] | 476 | querystring.clear();
|
---|
| 477 |
|
---|
[12784] | 478 | const int argct = args.getintarg("ct");
|
---|
| 479 | int argt = 0;// arg t is either not used (lucene) or used for natural/ranked (mgpp), so we set it to 0 = AND, by default
|
---|
| 480 | int argb = args.getintarg("b");
|
---|
[8029] | 481 | text_t combine;
|
---|
[12784] | 482 | if (argct==1) {
|
---|
[8029] | 483 | combine = "&";
|
---|
| 484 | }
|
---|
| 485 | else { // lucene
|
---|
| 486 | combine = "AND";
|
---|
| 487 | }
|
---|
| 488 |
|
---|
[1914] | 489 | text_t field = args["fqf"];
|
---|
| 490 | if (field.empty()) return; // no query
|
---|
| 491 | text_tarray fields;
|
---|
| 492 | splitchar(field.begin(), field.end(), ',', fields);
|
---|
| 493 |
|
---|
| 494 | text_t value = args["fqv"];
|
---|
| 495 | if (value.empty()) return; // somethings wrong
|
---|
| 496 | text_tarray values;
|
---|
| 497 | splitchar(value.begin(), value.end(), ',', values);
|
---|
| 498 |
|
---|
| 499 | text_t comb = args["fqc"];
|
---|
| 500 | if (comb.empty()) return; //somethings wrong
|
---|
| 501 | text_tarray combs;
|
---|
| 502 | splitchar(comb.begin(), comb.end(), ',', combs);
|
---|
[12784] | 503 |
|
---|
| 504 | text_tarray stems;
|
---|
| 505 | text_tarray folds;
|
---|
| 506 | if (argct == 1) {// mgpp - lucene doesn't do stem/case
|
---|
| 507 | text_t stem = args["fqs"];
|
---|
| 508 | if (stem.empty()) return; // somethings wrong
|
---|
| 509 | splitchar(stem.begin(), stem.end(), ',', stems);
|
---|
| 510 |
|
---|
| 511 | text_t fold = args["fqk"];
|
---|
| 512 | if (fold.empty()) return; // somethings wrong
|
---|
| 513 | splitchar(fold.begin(), fold.end(), ',', folds);
|
---|
| 514 | }
|
---|
[1914] | 515 |
|
---|
[9620] | 516 | for(int i=0; i< values.size(); ++i) {
|
---|
[1914] | 517 | if (!values[i].empty()) {
|
---|
| 518 | if (i!=0) {
|
---|
[12784] | 519 | if (argct==1) {
|
---|
[8029] | 520 | if (combs[i-1]=="and") combine = "&";
|
---|
| 521 | else if (combs[i-1]=="or")combine = "|";
|
---|
| 522 | else if (combs[i-1]=="not")combine = "!";
|
---|
| 523 | }
|
---|
| 524 | else { // lucene
|
---|
| 525 | if (combs[i-1]=="and") combine = "AND";
|
---|
| 526 | else if (combs[i-1]=="or")combine = "OR";
|
---|
| 527 | else if (combs[i-1]=="not")combine = "NOT";
|
---|
| 528 | }
|
---|
[1914] | 529 | }
|
---|
[12784] | 530 | text_t this_value = values[i];
|
---|
| 531 | // remove operators for simple search, segments text if necessary
|
---|
| 532 | format_querystring(this_value, argb, segment);
|
---|
| 533 | if (argct == 1) { // mgpp only
|
---|
| 534 | this_value = addstemcase(this_value, stems[i], folds[i], argct);
|
---|
[1914] | 535 | }
|
---|
[12784] | 536 | // add tag info for this field (and other processing)
|
---|
| 537 | format_field_info(this_value, fields[i], argct, argt, argb);
|
---|
| 538 | // add into query string
|
---|
| 539 | if (!querystring.empty()) {
|
---|
| 540 | querystring += " "+ combine+ " ";
|
---|
[2745] | 541 | }
|
---|
[12784] | 542 | querystring += this_value;
|
---|
[1914] | 543 |
|
---|
| 544 | }
|
---|
| 545 | }
|
---|
| 546 | }
|
---|
| 547 |
|
---|
[12784] | 548 | // Extended addqueryelem for Human Info project
|
---|
[7380] | 549 | void addqueryelem_ex(text_t &querystring, const text_t &tag,
|
---|
[12784] | 550 | const text_t &terms, const text_t &stem,
|
---|
| 551 | const text_t &fold,
|
---|
[7380] | 552 | const text_t& combine, const text_t& word_combine) {
|
---|
[12784] | 553 |
|
---|
[7380] | 554 | if (!querystring.empty()) { // have to put and/or
|
---|
| 555 | querystring += " " + combine + " ";
|
---|
| 556 | }
|
---|
| 557 | text_t outtext; outtext.reserve(512);
|
---|
| 558 | text_t word; word.reserve(100);
|
---|
| 559 | //unsigned short c;
|
---|
| 560 | text_t::const_iterator here = terms.begin();
|
---|
| 561 | text_t::const_iterator end = terms.end();
|
---|
| 562 | bool inquote = false, firstword = true;
|
---|
[1914] | 563 |
|
---|
[7380] | 564 | text_t word2; word2.reserve(256);
|
---|
| 565 |
|
---|
| 566 | while (here !=end) {
|
---|
| 567 | if (is_unicode_space(*here)) {
|
---|
| 568 | if (word2 == "AND") { word2.clear(); word2.push_back(7527); word2.appendcarr("AND", 3); word2.push_back(7527); }
|
---|
| 569 | else if (word2 == "OR") { word2.clear(); word2.push_back(7527); word2.appendcarr("OR", 2); word2.push_back(7527); }
|
---|
| 570 | else if (word2 == "NOT") { word2.clear(); word2.push_back(7527); word2.appendcarr("NOT", 3); word2.push_back(7527); }
|
---|
| 571 | else if (word2 == "NEAR") { word2.clear(); word2.push_back(7527); word2.appendcarr("NEAR", 4); word2.push_back(7527); }
|
---|
| 572 | else if (word2 == "WITHIN") { word2.clear(); word2.push_back(7527); word2.appendcarr("WITHIN", 6); word2.push_back(7527); }
|
---|
| 573 | if (inquote) {
|
---|
| 574 | word2.push_back(*here);
|
---|
| 575 | }
|
---|
| 576 | word.append(word2); word2.clear();
|
---|
| 577 |
|
---|
| 578 | if (!inquote && !word.empty() ) {
|
---|
[12784] | 579 | // found word boundary
|
---|
[7380] | 580 |
|
---|
| 581 | if (stem == "1" || fold =="1") {
|
---|
| 582 | word += "#";
|
---|
| 583 | if (stem == "1") word += "s";
|
---|
| 584 | //else word += "u";
|
---|
| 585 |
|
---|
| 586 | if (fold == "1") word += "i";
|
---|
| 587 | //else word += "c";
|
---|
| 588 | }
|
---|
| 589 | if (firstword) {
|
---|
| 590 | firstword = false;
|
---|
| 591 | } else {
|
---|
| 592 | outtext += " " + word_combine + " ";
|
---|
| 593 | }
|
---|
| 594 | outtext += "[" + word + "]:"+tag;
|
---|
| 595 | word.clear();
|
---|
| 596 | }
|
---|
| 597 | ++here;
|
---|
| 598 | } else if (*here == '\"') {
|
---|
| 599 | word2.push_back(*here);
|
---|
| 600 | inquote = !inquote;
|
---|
| 601 | ++here;
|
---|
| 602 | } else {
|
---|
| 603 | // not word boundary
|
---|
| 604 | word2.push_back(*here);
|
---|
| 605 | ++here;
|
---|
| 606 | }
|
---|
| 607 | }
|
---|
| 608 |
|
---|
| 609 | // get last word
|
---|
| 610 | if (!word2.empty()) {
|
---|
| 611 | if (word2 == "AND") { word2.clear(); word2.push_back(7527); word2.appendcarr("AND", 3); word2.push_back(7527); }
|
---|
| 612 | else if (word2 == "OR") { word2.clear(); word2.push_back(7527); word2.appendcarr("OR", 2); word2.push_back(7527); }
|
---|
| 613 | else if (word2 == "NOT") { word2.clear(); word2.push_back(7527); word2.appendcarr("NOT", 3); word2.push_back(7527); }
|
---|
| 614 | else if (word2 == "NEAR") { word2.clear(); word2.push_back(7527); word2.appendcarr("NEAR", 4); word2.push_back(7527); }
|
---|
| 615 | else if (word2 == "WITHIN") { word2.clear(); word2.push_back(7527); word2.appendcarr("WITHIN", 6); word2.push_back(7527); }
|
---|
| 616 | word.append(word2); word2.clear();
|
---|
| 617 |
|
---|
| 618 | if (stem == "1"|| fold == "1") {
|
---|
| 619 | word += "#";
|
---|
| 620 | if (stem == "1") word += "s";
|
---|
| 621 | //else word += "u";
|
---|
| 622 |
|
---|
| 623 | if (fold == "1") word += "i";
|
---|
| 624 | //else word += "c";
|
---|
| 625 | }
|
---|
| 626 | if (!outtext.empty()) outtext += " " + word_combine + " ";
|
---|
| 627 | outtext += "[" + word + "]:"+tag;
|
---|
| 628 | }
|
---|
| 629 | querystring += "(" + outtext + ")";
|
---|
| 630 | }
|
---|
| 631 |
|
---|
[8357] | 632 | void add_field_info(text_t &querystring, const text_t &tag, int type) {
|
---|
[7380] | 633 |
|
---|
[17796] | 634 | if (tag == "") return; // do nothing
|
---|
| 635 | if (tag == "ZZ" && type == 1) return; // mgpp doesn't use ZZ tag internally
|
---|
[8357] | 636 | if (type == 1) { //mgpp
|
---|
| 637 | querystring = "["+querystring+"]:"+tag;
|
---|
| 638 | } else if (type == 2) { // lucene
|
---|
| 639 | querystring = tag+":("+querystring+")";
|
---|
[4757] | 640 | }
|
---|
[8357] | 641 |
|
---|
[4757] | 642 | }
|
---|
[8029] | 643 |
|
---|
| 644 |
|
---|
[17796] | 645 | void format_field_info_lucene(text_t &querystring, text_t &tag, int argt, int argb) {
|
---|
| 646 |
|
---|
[11765] | 647 | int type = 2; //lucene
|
---|
[8029] | 648 |
|
---|
[12784] | 649 | if (argb==0) { // simple
|
---|
| 650 | // there will be no & or | as they should have already been removed
|
---|
[11765] | 651 | // just tag the entire thing
|
---|
[10995] | 652 | if (tag != "") {
|
---|
[11765] | 653 | add_field_info(querystring, tag, type);
|
---|
[10995] | 654 | }
|
---|
[8357] | 655 | return;
|
---|
| 656 | }
|
---|
[10995] | 657 |
|
---|
[12784] | 658 | // need to replace & with &&, | with ||
|
---|
[8357] | 659 | text_t::const_iterator here = querystring.begin();
|
---|
| 660 | text_t::const_iterator end = querystring.end();
|
---|
[12784] | 661 |
|
---|
| 662 | text_t finalquery = "";
|
---|
[10995] | 663 | while (here != end) {
|
---|
[12784] | 664 | if (*here == '&') {
|
---|
| 665 | finalquery.push_back('&');
|
---|
| 666 | finalquery.push_back('&');
|
---|
| 667 | while (*(here+1) == '&') {
|
---|
| 668 | ++here;
|
---|
[10995] | 669 | }
|
---|
[12784] | 670 | }
|
---|
| 671 | else if (*here == '|') {
|
---|
| 672 | finalquery.push_back('|');
|
---|
| 673 | finalquery.push_back('|');
|
---|
| 674 | while (*(here+1) == '|') {
|
---|
| 675 | ++here;
|
---|
| 676 | }
|
---|
| 677 | }
|
---|
[8357] | 678 | else {
|
---|
[12784] | 679 | finalquery.push_back(*here);
|
---|
[8357] | 680 | }
|
---|
[10995] | 681 | ++here;
|
---|
[8357] | 682 | }
|
---|
[11765] | 683 | querystring = finalquery;
|
---|
[12784] | 684 | add_field_info(querystring, tag, type);
|
---|
[11765] | 685 | }
|
---|
| 686 |
|
---|
[12784] | 687 |
|
---|
| 688 | void format_field_info_mgpp(text_t &querystring, text_t tag, int argt, int argb) {
|
---|
| 689 |
|
---|
[11765] | 690 | if (tag == "ZZ") tag = ""; // ZZ is a special tag meaning no tag (all fields)
|
---|
[12784] | 691 | if (tag == "" && argb == 1) {
|
---|
[11765] | 692 | return; // no field specifier, advanced mode, the query stays as written
|
---|
[10995] | 693 | }
|
---|
[11765] | 694 |
|
---|
| 695 | int type = 1; // mgpp
|
---|
| 696 |
|
---|
| 697 | bool simple_and = (argb==0 && argt==0);
|
---|
| 698 | text_t finalquery = "";
|
---|
| 699 | text_t fieldpart ="";
|
---|
| 700 | text_t queryelem = "";
|
---|
| 701 | bool in_phrase = false;
|
---|
| 702 | bool in_field = false;
|
---|
| 703 |
|
---|
| 704 | text_t::const_iterator here = querystring.begin();
|
---|
| 705 | text_t::const_iterator end = querystring.end();
|
---|
| 706 | while (here != end) {
|
---|
| 707 | if (is_unicode_letdig(*here) || *here == '&' || is_special_character(type, *here)) {
|
---|
| 708 | queryelem.push_back(*here);
|
---|
| 709 | }
|
---|
| 710 | else if (*here == '|') {
|
---|
| 711 | in_field = false;
|
---|
| 712 | }
|
---|
| 713 | else if (*here == '!' || *here == '(' || *here == ')') {
|
---|
| 714 | if (!in_phrase) { // ignore these if in_phrase
|
---|
| 715 | // output field, then output operator
|
---|
| 716 | in_field = false;
|
---|
| 717 | if (!queryelem.empty()) {
|
---|
| 718 | if (!simple_and && !fieldpart.empty()) {
|
---|
| 719 | add_field_info(fieldpart, tag, type);
|
---|
| 720 | finalquery += fieldpart;
|
---|
| 721 | finalquery.push_back(' ');
|
---|
| 722 | fieldpart.clear();
|
---|
| 723 | }
|
---|
| 724 | fieldpart += queryelem;
|
---|
| 725 | }
|
---|
| 726 | if (!fieldpart.empty()) {
|
---|
| 727 | add_field_info(fieldpart, tag, type);
|
---|
| 728 | finalquery += fieldpart;
|
---|
| 729 | finalquery.push_back(' ');
|
---|
| 730 | }
|
---|
| 731 | fieldpart.clear();
|
---|
| 732 | queryelem.clear();
|
---|
| 733 | finalquery.push_back(*here);
|
---|
| 734 | finalquery.push_back(' ');
|
---|
| 735 | }
|
---|
| 736 | }
|
---|
| 737 | else if (*here == '"') {
|
---|
| 738 | queryelem.push_back(*here);
|
---|
| 739 | if (in_phrase == false) in_phrase = true;
|
---|
| 740 | else {
|
---|
| 741 | in_phrase = false;
|
---|
| 742 | }
|
---|
| 743 | }
|
---|
| 744 |
|
---|
| 745 | // Found word boundary, in a phrase
|
---|
| 746 | else if (in_phrase) {
|
---|
| 747 | queryelem.push_back(*here);
|
---|
| 748 | }
|
---|
| 749 | // Found a word boundary
|
---|
| 750 | else {
|
---|
| 751 | if (!queryelem.empty()) {
|
---|
| 752 | if (queryelem == "&") {
|
---|
| 753 | in_field = true;
|
---|
| 754 | queryelem.clear();
|
---|
| 755 | }
|
---|
| 756 | else if (starts_with(queryelem, "NEAR") || starts_with(queryelem, "WITHIN")) {
|
---|
| 757 |
|
---|
| 758 | if (argb==1) {
|
---|
| 759 | // simple search, these not allowed
|
---|
| 760 | in_field = true;
|
---|
| 761 | fieldpart += queryelem;
|
---|
| 762 | fieldpart.push_back(' ');
|
---|
| 763 | }
|
---|
| 764 | queryelem.clear();
|
---|
| 765 |
|
---|
| 766 | }
|
---|
| 767 | else {
|
---|
| 768 | if (!simple_and && !in_field) {
|
---|
| 769 | if (!fieldpart.empty()) {
|
---|
| 770 | add_field_info(fieldpart, tag, type);
|
---|
| 771 | finalquery += fieldpart;
|
---|
| 772 | finalquery.push_back(' ');
|
---|
| 773 | fieldpart.clear();
|
---|
| 774 | }
|
---|
| 775 | }
|
---|
| 776 |
|
---|
| 777 | fieldpart += queryelem;
|
---|
| 778 | fieldpart.push_back(' ');
|
---|
| 779 | queryelem.clear();
|
---|
| 780 | }
|
---|
| 781 | }
|
---|
| 782 | }
|
---|
| 783 | ++here;
|
---|
| 784 | }
|
---|
| 785 | // at the end
|
---|
| 786 | if (!queryelem.empty()) {
|
---|
| 787 | if (!simple_and && !in_field && !fieldpart.empty()) {
|
---|
| 788 | add_field_info(fieldpart, tag, type);
|
---|
| 789 | finalquery += fieldpart;
|
---|
| 790 | finalquery.push_back(' ');
|
---|
| 791 | fieldpart.clear();
|
---|
| 792 | }
|
---|
| 793 | fieldpart += queryelem;
|
---|
| 794 | }
|
---|
| 795 | if (!fieldpart.empty()) {
|
---|
| 796 | add_field_info(fieldpart, tag, type);
|
---|
| 797 | finalquery += fieldpart;
|
---|
| 798 | fieldpart.clear();
|
---|
| 799 | finalquery.push_back(' ');
|
---|
| 800 | }
|
---|
| 801 |
|
---|
| 802 | querystring = finalquery;
|
---|
[8029] | 803 | }
|
---|
[8357] | 804 |
|
---|
[12784] | 805 |
|
---|
| 806 | void format_field_info(text_t &querystring, text_t tag, int argct, int argt, int argb) {
|
---|
[11765] | 807 | if (argct == 1) {
|
---|
[12784] | 808 | format_field_info_mgpp(querystring, tag, argt, argb);
|
---|
[11765] | 809 | } else if (argct == 2) {
|
---|
[12784] | 810 | format_field_info_lucene(querystring, tag, argt, argb);
|
---|
[11765] | 811 | }
|
---|
| 812 | }
|
---|
[10995] | 813 |
|
---|
[12784] | 814 | void mgpp_adddateelem(text_t& querystring, const int date)
|
---|
| 815 | {
|
---|
| 816 | querystring.appendcstr(" [");
|
---|
| 817 | if(date<0) {
|
---|
| 818 | querystring.appendcstr("bc");
|
---|
| 819 | querystring.appendint((date*-1));
|
---|
| 820 | }
|
---|
| 821 | else {
|
---|
| 822 | querystring.appendint(date);
|
---|
| 823 | }
|
---|
| 824 | querystring.appendcstr("]:CV");
|
---|
| 825 | }
|
---|
| 826 |
|
---|
| 827 | void lucene_adddateelem(text_t& querystring, const int date)
|
---|
| 828 | {
|
---|
| 829 | querystring.appendcstr(" CV:(");
|
---|
| 830 | if(date<0) {
|
---|
| 831 | querystring.appendcstr("bc");
|
---|
| 832 | querystring.appendint((date*-1));
|
---|
| 833 | }
|
---|
| 834 | else {
|
---|
| 835 | querystring.appendint(date);
|
---|
| 836 | }
|
---|
| 837 | querystring.appendcstr(")");
|
---|
| 838 | }
|
---|
| 839 |
|
---|
| 840 |
|
---|
| 841 | void add_dates(text_t &querystring, int startdate, int enddate,
|
---|
| 842 | int startbc, int endbc, int ct)
|
---|
| 843 | {
|
---|
| 844 | if(startdate)
|
---|
| 845 | {
|
---|
| 846 | int querystringis = 0;
|
---|
| 847 | text_t::const_iterator here = querystring.begin();
|
---|
| 848 | text_t::const_iterator end = querystring.end();
|
---|
| 849 | while(here!=end)
|
---|
| 850 | {
|
---|
| 851 | if(!(isspace((*here)))){
|
---|
| 852 | here = end;
|
---|
| 853 | querystringis = 1;
|
---|
| 854 | }
|
---|
| 855 | else
|
---|
| 856 | ++here;
|
---|
| 857 | }
|
---|
| 858 | //converting BCE dates
|
---|
| 859 | if(startbc && startdate > 0)
|
---|
| 860 | {
|
---|
| 861 | startdate *= -1;
|
---|
| 862 | }
|
---|
| 863 | if(endbc && enddate > 0)
|
---|
| 864 | {
|
---|
| 865 | enddate *= -1;
|
---|
| 866 | }
|
---|
| 867 | if(enddate != 0 && enddate<startdate)
|
---|
| 868 | {
|
---|
| 869 | cout<<"enddate too small"<<endl;
|
---|
| 870 | return;
|
---|
| 871 | }
|
---|
| 872 | if(querystringis)
|
---|
| 873 | querystring.appendcstr(" AND");
|
---|
| 874 | if(!enddate)
|
---|
| 875 | {
|
---|
| 876 | if (ct==1) {
|
---|
| 877 | mgpp_adddateelem(querystring,startdate);
|
---|
| 878 | }
|
---|
| 879 | else { // lucene
|
---|
| 880 | lucene_adddateelem(querystring,startdate);
|
---|
| 881 | }
|
---|
| 882 | }
|
---|
| 883 | else{
|
---|
| 884 | int nextdate = startdate;
|
---|
| 885 | querystring.appendcstr(" (");
|
---|
| 886 | while(nextdate<=enddate)
|
---|
| 887 | {
|
---|
| 888 | if(nextdate!=0) {
|
---|
| 889 | if (ct==1) {
|
---|
| 890 | mgpp_adddateelem(querystring,nextdate);
|
---|
| 891 | }
|
---|
| 892 | else { // lucene
|
---|
| 893 | lucene_adddateelem(querystring,nextdate);
|
---|
| 894 | }
|
---|
| 895 | }
|
---|
| 896 | ++nextdate;
|
---|
| 897 | }
|
---|
| 898 | querystring.appendcstr(" )");
|
---|
| 899 | }
|
---|
| 900 | }
|
---|
| 901 |
|
---|
| 902 | }
|
---|