[1113] | 1 | <?php |
---|
| 2 | // +-----------------------------------------------------------------------+ |
---|
[8728] | 3 | // | Piwigo - a PHP based photo gallery | |
---|
[2297] | 4 | // +-----------------------------------------------------------------------+ |
---|
[19703] | 5 | // | Copyright(C) 2008-2013 Piwigo Team http://piwigo.org | |
---|
[2297] | 6 | // | Copyright(C) 2003-2008 PhpWebGallery Team http://phpwebgallery.net | |
---|
| 7 | // | Copyright(C) 2002-2003 Pierrick LE GALL http://le-gall.net/pierrick | |
---|
| 8 | // +-----------------------------------------------------------------------+ |
---|
| 9 | // | This program is free software; you can redistribute it and/or modify | |
---|
| 10 | // | it under the terms of the GNU General Public License as published by | |
---|
| 11 | // | the Free Software Foundation | |
---|
| 12 | // | | |
---|
| 13 | // | This program is distributed in the hope that it will be useful, but | |
---|
| 14 | // | WITHOUT ANY WARRANTY; without even the implied warranty of | |
---|
| 15 | // | MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU | |
---|
| 16 | // | General Public License for more details. | |
---|
| 17 | // | | |
---|
| 18 | // | You should have received a copy of the GNU General Public License | |
---|
| 19 | // | along with this program; if not, write to the Free Software | |
---|
| 20 | // | Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, | |
---|
| 21 | // | USA. | |
---|
| 22 | // +-----------------------------------------------------------------------+ |
---|
[1113] | 23 | |
---|
[25658] | 24 | /** |
---|
| 25 | * @package functions\search |
---|
| 26 | */ |
---|
[1113] | 27 | |
---|
[25658] | 28 | |
---|
[1113] | 29 | /** |
---|
[25658] | 30 | * Returns search rules stored into a serialized array in "search" |
---|
[1113] | 31 | * table. Each search rules set is numericaly identified. |
---|
| 32 | * |
---|
[25658] | 33 | * @param int $search_id |
---|
[1113] | 34 | * @return array |
---|
| 35 | */ |
---|
| 36 | function get_search_array($search_id) |
---|
| 37 | { |
---|
| 38 | if (!is_numeric($search_id)) |
---|
| 39 | { |
---|
| 40 | die('Search id must be an integer'); |
---|
| 41 | } |
---|
| 42 | |
---|
| 43 | $query = ' |
---|
| 44 | SELECT rules |
---|
| 45 | FROM '.SEARCH_TABLE.' |
---|
| 46 | WHERE id = '.$search_id.' |
---|
| 47 | ;'; |
---|
[4325] | 48 | list($serialized_rules) = pwg_db_fetch_row(pwg_query($query)); |
---|
[1113] | 49 | |
---|
| 50 | return unserialize($serialized_rules); |
---|
| 51 | } |
---|
| 52 | |
---|
| 53 | /** |
---|
[25658] | 54 | * Returns the SQL clause for a search. |
---|
| 55 | * Transforms the array returned by get_search_array() into SQL sub-query. |
---|
[1113] | 56 | * |
---|
[25658] | 57 | * @param array $search |
---|
[1113] | 58 | * @return string |
---|
| 59 | */ |
---|
[1537] | 60 | function get_sql_search_clause($search) |
---|
[1113] | 61 | { |
---|
| 62 | // SQL where clauses are stored in $clauses array during query |
---|
| 63 | // construction |
---|
| 64 | $clauses = array(); |
---|
| 65 | |
---|
[1119] | 66 | foreach (array('file','name','comment','author') as $textfield) |
---|
[1113] | 67 | { |
---|
| 68 | if (isset($search['fields'][$textfield])) |
---|
| 69 | { |
---|
| 70 | $local_clauses = array(); |
---|
| 71 | foreach ($search['fields'][$textfield]['words'] as $word) |
---|
| 72 | { |
---|
[25018] | 73 | $local_clauses[] = $textfield." LIKE '%".$word."%'"; |
---|
[1113] | 74 | } |
---|
| 75 | |
---|
| 76 | // adds brackets around where clauses |
---|
| 77 | $local_clauses = prepend_append_array_items($local_clauses, '(', ')'); |
---|
| 78 | |
---|
[25018] | 79 | $clauses[] = implode( |
---|
| 80 | ' '.$search['fields'][$textfield]['mode'].' ', |
---|
| 81 | $local_clauses |
---|
[1113] | 82 | ); |
---|
| 83 | } |
---|
| 84 | } |
---|
| 85 | |
---|
| 86 | if (isset($search['fields']['allwords'])) |
---|
| 87 | { |
---|
[1119] | 88 | $fields = array('file', 'name', 'comment', 'author'); |
---|
[1113] | 89 | // in the OR mode, request bust be : |
---|
| 90 | // ((field1 LIKE '%word1%' OR field2 LIKE '%word1%') |
---|
| 91 | // OR (field1 LIKE '%word2%' OR field2 LIKE '%word2%')) |
---|
| 92 | // |
---|
| 93 | // in the AND mode : |
---|
| 94 | // ((field1 LIKE '%word1%' OR field2 LIKE '%word1%') |
---|
| 95 | // AND (field1 LIKE '%word2%' OR field2 LIKE '%word2%')) |
---|
| 96 | $word_clauses = array(); |
---|
| 97 | foreach ($search['fields']['allwords']['words'] as $word) |
---|
| 98 | { |
---|
| 99 | $field_clauses = array(); |
---|
| 100 | foreach ($fields as $field) |
---|
| 101 | { |
---|
[25018] | 102 | $field_clauses[] = $field." LIKE '%".$word."%'"; |
---|
[1113] | 103 | } |
---|
| 104 | // adds brackets around where clauses |
---|
[25018] | 105 | $word_clauses[] = implode( |
---|
| 106 | "\n OR ", |
---|
| 107 | $field_clauses |
---|
[1113] | 108 | ); |
---|
| 109 | } |
---|
| 110 | |
---|
| 111 | array_walk( |
---|
| 112 | $word_clauses, |
---|
| 113 | create_function('&$s','$s="(".$s.")";') |
---|
| 114 | ); |
---|
| 115 | |
---|
[25018] | 116 | $clauses[] = "\n ". |
---|
[1113] | 117 | implode( |
---|
[25018] | 118 | "\n ". $search['fields']['allwords']['mode']. "\n ", |
---|
[1113] | 119 | $word_clauses |
---|
[25018] | 120 | ); |
---|
[1113] | 121 | } |
---|
| 122 | |
---|
| 123 | foreach (array('date_available', 'date_creation') as $datefield) |
---|
| 124 | { |
---|
| 125 | if (isset($search['fields'][$datefield])) |
---|
| 126 | { |
---|
[25026] | 127 | $clauses[] = $datefield." = '".$search['fields'][$datefield]['date']."'"; |
---|
[1113] | 128 | } |
---|
| 129 | |
---|
| 130 | foreach (array('after','before') as $suffix) |
---|
| 131 | { |
---|
| 132 | $key = $datefield.'-'.$suffix; |
---|
| 133 | |
---|
| 134 | if (isset($search['fields'][$key])) |
---|
| 135 | { |
---|
[25018] | 136 | $clauses[] = $datefield. |
---|
[1113] | 137 | ($suffix == 'after' ? ' >' : ' <'). |
---|
| 138 | ($search['fields'][$key]['inc'] ? '=' : ''). |
---|
[25018] | 139 | " '".$search['fields'][$key]['date']."'"; |
---|
[1113] | 140 | } |
---|
| 141 | } |
---|
| 142 | } |
---|
| 143 | |
---|
| 144 | if (isset($search['fields']['cat'])) |
---|
| 145 | { |
---|
| 146 | if ($search['fields']['cat']['sub_inc']) |
---|
| 147 | { |
---|
| 148 | // searching all the categories id of sub-categories |
---|
| 149 | $cat_ids = get_subcat_ids($search['fields']['cat']['words']); |
---|
| 150 | } |
---|
| 151 | else |
---|
| 152 | { |
---|
| 153 | $cat_ids = $search['fields']['cat']['words']; |
---|
| 154 | } |
---|
| 155 | |
---|
| 156 | $local_clause = 'category_id IN ('.implode(',', $cat_ids).')'; |
---|
[25018] | 157 | $clauses[] = $local_clause; |
---|
[1113] | 158 | } |
---|
| 159 | |
---|
| 160 | // adds brackets around where clauses |
---|
| 161 | $clauses = prepend_append_array_items($clauses, '(', ')'); |
---|
| 162 | |
---|
| 163 | $where_separator = |
---|
| 164 | implode( |
---|
| 165 | "\n ".$search['mode'].' ', |
---|
| 166 | $clauses |
---|
| 167 | ); |
---|
| 168 | |
---|
| 169 | $search_clause = $where_separator; |
---|
| 170 | |
---|
[1119] | 171 | return $search_clause; |
---|
| 172 | } |
---|
| 173 | |
---|
| 174 | /** |
---|
[25658] | 175 | * Returns the list of items corresponding to the advanced search array. |
---|
[1119] | 176 | * |
---|
[25658] | 177 | * @param array $search |
---|
| 178 | * @param string $images_where optional additional restriction on images table |
---|
[1119] | 179 | * @return array |
---|
| 180 | */ |
---|
[25658] | 181 | function get_regular_search_results($search, $images_where='') |
---|
[1119] | 182 | { |
---|
[2451] | 183 | global $conf; |
---|
| 184 | $forbidden = get_sql_condition_FandF( |
---|
| 185 | array |
---|
| 186 | ( |
---|
| 187 | 'forbidden_categories' => 'category_id', |
---|
| 188 | 'visible_categories' => 'category_id', |
---|
| 189 | 'visible_images' => 'id' |
---|
| 190 | ), |
---|
| 191 | "\n AND" |
---|
| 192 | ); |
---|
| 193 | |
---|
[1119] | 194 | $items = array(); |
---|
[2451] | 195 | $tag_items = array(); |
---|
[1537] | 196 | |
---|
[2451] | 197 | if (isset($search['fields']['tags'])) |
---|
| 198 | { |
---|
| 199 | $tag_items = get_image_ids_for_tags( |
---|
| 200 | $search['fields']['tags']['words'], |
---|
| 201 | $search['fields']['tags']['mode'] |
---|
| 202 | ); |
---|
| 203 | } |
---|
| 204 | |
---|
[1537] | 205 | $search_clause = get_sql_search_clause($search); |
---|
| 206 | |
---|
[1119] | 207 | if (!empty($search_clause)) |
---|
[1113] | 208 | { |
---|
[1119] | 209 | $query = ' |
---|
[8611] | 210 | SELECT DISTINCT(id) |
---|
[2451] | 211 | FROM '.IMAGES_TABLE.' i |
---|
[1119] | 212 | INNER JOIN '.IMAGE_CATEGORY_TABLE.' AS ic ON id = ic.image_id |
---|
[2451] | 213 | WHERE '.$search_clause; |
---|
| 214 | if (!empty($images_where)) |
---|
| 215 | { |
---|
| 216 | $query .= "\n AND ".$images_where; |
---|
| 217 | } |
---|
[8726] | 218 | $query .= $forbidden.' |
---|
[2451] | 219 | '.$conf['order_by']; |
---|
[1119] | 220 | $items = array_from_query($query, 'id'); |
---|
[1113] | 221 | } |
---|
| 222 | |
---|
[2451] | 223 | if ( !empty($tag_items) ) |
---|
[1119] | 224 | { |
---|
| 225 | switch ($search['mode']) |
---|
| 226 | { |
---|
| 227 | case 'AND': |
---|
| 228 | if (empty($search_clause)) |
---|
| 229 | { |
---|
| 230 | $items = $tag_items; |
---|
| 231 | } |
---|
| 232 | else |
---|
| 233 | { |
---|
[5691] | 234 | $items = array_values( array_intersect($items, $tag_items) ); |
---|
[1119] | 235 | } |
---|
| 236 | break; |
---|
| 237 | case 'OR': |
---|
[2451] | 238 | $before_count = count($items); |
---|
[1119] | 239 | $items = array_unique( |
---|
| 240 | array_merge( |
---|
| 241 | $items, |
---|
| 242 | $tag_items |
---|
| 243 | ) |
---|
| 244 | ); |
---|
| 245 | break; |
---|
[2451] | 246 | } |
---|
[1119] | 247 | } |
---|
[1537] | 248 | |
---|
[1119] | 249 | return $items; |
---|
[1113] | 250 | } |
---|
[1537] | 251 | |
---|
[25658] | 252 | /** |
---|
| 253 | * Finds if a char is a letter, a figure or any char of the extended ASCII table (>127). |
---|
| 254 | * |
---|
| 255 | * @param char $ch |
---|
| 256 | * @return bool |
---|
| 257 | */ |
---|
[10340] | 258 | function is_word_char($ch) |
---|
| 259 | { |
---|
| 260 | return ($ch>='0' && $ch<='9') || ($ch>='a' && $ch<='z') || ($ch>='A' && $ch<='Z') || ord($ch)>127; |
---|
| 261 | } |
---|
| 262 | |
---|
[25658] | 263 | /** |
---|
| 264 | * Finds if a char is a special token for word start: [{<=*+ |
---|
| 265 | * |
---|
| 266 | * @param char $ch |
---|
| 267 | * @return bool |
---|
| 268 | */ |
---|
[18207] | 269 | function is_odd_wbreak_begin($ch) |
---|
| 270 | { |
---|
| 271 | return strpos('[{<=*+', $ch)===false ? false:true; |
---|
| 272 | } |
---|
| 273 | |
---|
[25658] | 274 | /** |
---|
| 275 | * Finds if a char is a special token for word end: ]}>=*+ |
---|
| 276 | * |
---|
| 277 | * @param char $ch |
---|
| 278 | * @return bool |
---|
| 279 | */ |
---|
[18207] | 280 | function is_odd_wbreak_end($ch) |
---|
| 281 | { |
---|
| 282 | return strpos(']}>=*+', $ch)===false ? false:true; |
---|
| 283 | } |
---|
| 284 | |
---|
[25658] | 285 | |
---|
| 286 | define('QST_QUOTED', 0x01); |
---|
| 287 | define('QST_NOT', 0x02); |
---|
| 288 | define('QST_WILDCARD_BEGIN', 0x04); |
---|
| 289 | define('QST_WILDCARD_END', 0x08); |
---|
[18207] | 290 | define('QST_WILDCARD', QST_WILDCARD_BEGIN|QST_WILDCARD_END); |
---|
| 291 | |
---|
[1619] | 292 | /** |
---|
[25658] | 293 | * Analyzes and splits the quick/query search query $q into tokens. |
---|
[10340] | 294 | * q='john bill' => 2 tokens 'john' 'bill' |
---|
| 295 | * Special characters for MySql full text search (+,<,>,~) appear in the token modifiers. |
---|
| 296 | * The query can contain a phrase: 'Pierre "New York"' will return 'pierre' qnd 'new york'. |
---|
[25658] | 297 | * |
---|
| 298 | * @param string $q |
---|
| 299 | * @param array &$qtokens |
---|
| 300 | * @param array &$qtoken_modifiers |
---|
[1619] | 301 | */ |
---|
[10340] | 302 | function analyse_qsearch($q, &$qtokens, &$qtoken_modifiers) |
---|
[1537] | 303 | { |
---|
[2135] | 304 | $q = stripslashes($q); |
---|
| 305 | $tokens = array(); |
---|
| 306 | $token_modifiers = array(); |
---|
| 307 | $crt_token = ""; |
---|
[18207] | 308 | $crt_token_modifier = 0; |
---|
[2135] | 309 | |
---|
| 310 | for ($i=0; $i<strlen($q); $i++) |
---|
[1537] | 311 | { |
---|
[2135] | 312 | $ch = $q[$i]; |
---|
[18636] | 313 | if ( ($crt_token_modifier&QST_QUOTED)==0) |
---|
[1537] | 314 | { |
---|
[2135] | 315 | if ($ch=='"') |
---|
| 316 | { |
---|
[18207] | 317 | if (strlen($crt_token)) |
---|
| 318 | { |
---|
| 319 | $tokens[] = $crt_token; $token_modifiers[] = $crt_token_modifier; |
---|
| 320 | $crt_token = ""; $crt_token_modifier = 0; |
---|
| 321 | } |
---|
| 322 | $crt_token_modifier |= QST_QUOTED; |
---|
[2135] | 323 | } |
---|
[18207] | 324 | elseif ( strcspn($ch, '*+-><~')==0 ) |
---|
| 325 | { //special full text modifier |
---|
[10340] | 326 | if (strlen($crt_token)) |
---|
| 327 | { |
---|
| 328 | $crt_token .= $ch; |
---|
| 329 | } |
---|
| 330 | else |
---|
| 331 | { |
---|
[18207] | 332 | if ( $ch=='*' ) |
---|
| 333 | $crt_token_modifier |= QST_WILDCARD_BEGIN; |
---|
| 334 | if ( $ch=='-' ) |
---|
| 335 | $crt_token_modifier |= QST_NOT; |
---|
[10340] | 336 | } |
---|
[2135] | 337 | } |
---|
| 338 | elseif (preg_match('/[\s,.;!\?]+/', $ch)) |
---|
| 339 | { // white space |
---|
| 340 | if (strlen($crt_token)) |
---|
| 341 | { |
---|
[10340] | 342 | $tokens[] = $crt_token; $token_modifiers[] = $crt_token_modifier; |
---|
[18207] | 343 | $crt_token = ""; |
---|
[2135] | 344 | } |
---|
[18207] | 345 | $crt_token_modifier = 0; |
---|
[2135] | 346 | } |
---|
| 347 | else |
---|
| 348 | { |
---|
| 349 | $crt_token .= $ch; |
---|
| 350 | } |
---|
[18207] | 351 | } |
---|
| 352 | else // qualified with quotes |
---|
| 353 | { |
---|
| 354 | if ($ch=='"') |
---|
| 355 | { |
---|
| 356 | if ($i+1 < strlen($q) && $q[$i+1]=='*') |
---|
[2135] | 357 | { |
---|
[18207] | 358 | $crt_token_modifier |= QST_WILDCARD_END; |
---|
| 359 | $i++; |
---|
[2135] | 360 | } |
---|
[18207] | 361 | $tokens[] = $crt_token; $token_modifiers[] = $crt_token_modifier; |
---|
| 362 | $crt_token = ""; $crt_token_modifier = 0; |
---|
| 363 | $state=0; |
---|
| 364 | } |
---|
| 365 | else |
---|
| 366 | $crt_token .= $ch; |
---|
[1537] | 367 | } |
---|
| 368 | } |
---|
[18636] | 369 | |
---|
[2135] | 370 | if (strlen($crt_token)) |
---|
| 371 | { |
---|
| 372 | $tokens[] = $crt_token; |
---|
| 373 | $token_modifiers[] = $crt_token_modifier; |
---|
| 374 | } |
---|
[1537] | 375 | |
---|
[10340] | 376 | $qtokens = array(); |
---|
| 377 | $qtoken_modifiers = array(); |
---|
| 378 | for ($i=0; $i<count($tokens); $i++) |
---|
| 379 | { |
---|
[22175] | 380 | if ( !($token_modifiers[$i] & QST_QUOTED) ) |
---|
[10340] | 381 | { |
---|
| 382 | if ( substr($tokens[$i], -1)=='*' ) |
---|
| 383 | { |
---|
| 384 | $tokens[$i] = rtrim($tokens[$i], '*'); |
---|
[22175] | 385 | $token_modifiers[$i] |= QST_WILDCARD_END; |
---|
[10340] | 386 | } |
---|
| 387 | } |
---|
| 388 | if ( strlen($tokens[$i])==0) |
---|
| 389 | continue; |
---|
| 390 | $qtokens[] = $tokens[$i]; |
---|
| 391 | $qtoken_modifiers[] = $token_modifiers[$i]; |
---|
| 392 | } |
---|
| 393 | } |
---|
| 394 | |
---|
| 395 | /** |
---|
[25658] | 396 | * Returns the LIKE SQL clause corresponding to the quick search query |
---|
| 397 | * that has been split into tokens. |
---|
[10340] | 398 | * for example file LIKE '%john%' OR file LIKE '%bill%'. |
---|
[25658] | 399 | * |
---|
| 400 | * @param array $tokens |
---|
| 401 | * @param array $token_modifiers |
---|
| 402 | * @param string $field |
---|
| 403 | * @return string|null |
---|
[10340] | 404 | */ |
---|
| 405 | function get_qsearch_like_clause($tokens, $token_modifiers, $field) |
---|
| 406 | { |
---|
[2135] | 407 | $clauses = array(); |
---|
| 408 | for ($i=0; $i<count($tokens); $i++) |
---|
[1537] | 409 | { |
---|
[10340] | 410 | $token = trim($tokens[$i], '%'); |
---|
[18207] | 411 | if ($token_modifiers[$i]&QST_NOT) |
---|
[2135] | 412 | continue; |
---|
[11979] | 413 | if ( strlen($token)==0 ) |
---|
[2135] | 414 | continue; |
---|
[10340] | 415 | $token = addslashes($token); |
---|
| 416 | $token = str_replace( array('%','_'), array('\\%','\\_'), $token); // escape LIKE specials %_ |
---|
| 417 | $clauses[] = $field.' LIKE \'%'.$token.'%\''; |
---|
[1537] | 418 | } |
---|
[2135] | 419 | |
---|
| 420 | return count($clauses) ? '('.implode(' OR ', $clauses).')' : null; |
---|
[1537] | 421 | } |
---|
| 422 | |
---|
| 423 | /** |
---|
[25658] | 424 | * Returns tags corresponding to the quick search query that has been split into tokens. |
---|
| 425 | * |
---|
| 426 | * @param array $tokens |
---|
| 427 | * @param array $token_modifiers |
---|
| 428 | * @param array &$token_tag_ids |
---|
| 429 | * @param array &$not_tag_ids |
---|
| 430 | * @param array &$all_tags |
---|
| 431 | */ |
---|
[18636] | 432 | function get_qsearch_tags($tokens, $token_modifiers, &$token_tag_ids, &$not_tag_ids, &$all_tags) |
---|
[1537] | 433 | { |
---|
[18636] | 434 | $token_tag_ids = array_fill(0, count($tokens), array() ); |
---|
| 435 | $not_tag_ids = $all_tags = array(); |
---|
[10340] | 436 | |
---|
[18636] | 437 | $token_tag_scores = $token_tag_ids; |
---|
[10340] | 438 | $transliterated_tokens = array(); |
---|
| 439 | foreach ($tokens as $token) |
---|
| 440 | { |
---|
| 441 | $transliterated_tokens[] = transliterate($token); |
---|
| 442 | } |
---|
| 443 | |
---|
| 444 | $query = ' |
---|
[18636] | 445 | SELECT t.*, COUNT(image_id) AS counter |
---|
| 446 | FROM '.TAGS_TABLE.' t |
---|
[10340] | 447 | INNER JOIN '.IMAGE_TAG_TABLE.' ON id=tag_id |
---|
| 448 | GROUP BY id'; |
---|
| 449 | $result = pwg_query($query); |
---|
| 450 | while ($tag = pwg_db_fetch_assoc($result)) |
---|
| 451 | { |
---|
| 452 | $transliterated_tag = transliterate($tag['name']); |
---|
| 453 | |
---|
| 454 | // find how this tag matches query tokens |
---|
| 455 | for ($i=0; $i<count($tokens); $i++) |
---|
| 456 | { |
---|
| 457 | $transliterated_token = $transliterated_tokens[$i]; |
---|
| 458 | |
---|
| 459 | $match = false; |
---|
| 460 | $pos = 0; |
---|
| 461 | while ( ($pos = strpos($transliterated_tag, $transliterated_token, $pos)) !== false) |
---|
| 462 | { |
---|
[18207] | 463 | if ( ($token_modifiers[$i]&QST_WILDCARD)==QST_WILDCARD ) |
---|
[10340] | 464 | {// wildcard in this token |
---|
| 465 | $match = 1; |
---|
| 466 | break; |
---|
| 467 | } |
---|
| 468 | $token_len = strlen($transliterated_token); |
---|
| 469 | |
---|
[18207] | 470 | // search begin of word |
---|
| 471 | $wbegin_len=0; $wbegin_char=' '; |
---|
| 472 | while ($pos-$wbegin_len > 0) |
---|
[10340] | 473 | { |
---|
[18207] | 474 | if (! is_word_char($transliterated_tag[$pos-$wbegin_len-1]) ) |
---|
| 475 | { |
---|
| 476 | $wbegin_char = $transliterated_tag[$pos-$wbegin_len-1]; |
---|
[10340] | 477 | break; |
---|
[18207] | 478 | } |
---|
| 479 | $wbegin_len++; |
---|
[10340] | 480 | } |
---|
| 481 | |
---|
[18207] | 482 | // search end of word |
---|
| 483 | $wend_len=0; $wend_char=' '; |
---|
| 484 | while ($pos+$token_len+$wend_len < strlen($transliterated_tag)) |
---|
[10340] | 485 | { |
---|
[18207] | 486 | if (! is_word_char($transliterated_tag[$pos+$token_len+$wend_len]) ) |
---|
| 487 | { |
---|
| 488 | $wend_char = $transliterated_tag[$pos+$token_len+$wend_len]; |
---|
| 489 | break; |
---|
| 490 | } |
---|
| 491 | $wend_len++; |
---|
[10340] | 492 | } |
---|
| 493 | |
---|
[18207] | 494 | $this_score = 0; |
---|
| 495 | if ( ($token_modifiers[$i]&QST_WILDCARD)==0 ) |
---|
| 496 | {// no wildcard begin or end |
---|
| 497 | if ($token_len <= 2) |
---|
| 498 | {// search for 1 or 2 characters must match exactly to avoid retrieving too much data |
---|
| 499 | if ($wbegin_len==0 && $wend_len==0 && !is_odd_wbreak_begin($wbegin_char) && !is_odd_wbreak_end($wend_char) ) |
---|
| 500 | $this_score = 1; |
---|
| 501 | } |
---|
| 502 | elseif ($token_len == 3) |
---|
| 503 | { |
---|
| 504 | if ($wbegin_len==0) |
---|
| 505 | $this_score = $token_len / ($token_len + $wend_len); |
---|
| 506 | } |
---|
| 507 | else |
---|
| 508 | { |
---|
| 509 | $this_score = $token_len / ($token_len + 1.1 * $wbegin_len + 0.9 * $wend_len); |
---|
| 510 | } |
---|
| 511 | } |
---|
| 512 | |
---|
[10340] | 513 | if ($this_score>0) |
---|
| 514 | $match = max($match, $this_score ); |
---|
| 515 | $pos++; |
---|
| 516 | } |
---|
| 517 | |
---|
| 518 | if ($match) |
---|
| 519 | { |
---|
| 520 | $tag_id = (int)$tag['id']; |
---|
| 521 | $all_tags[$tag_id] = $tag; |
---|
[18636] | 522 | $token_tag_ids[$i][] = $tag_id; |
---|
| 523 | $token_tag_scores[$i][] = $match; |
---|
[10340] | 524 | } |
---|
| 525 | } |
---|
| 526 | } |
---|
| 527 | |
---|
[18636] | 528 | // process not tags |
---|
| 529 | for ($i=0; $i<count($tokens); $i++) |
---|
[10340] | 530 | { |
---|
[18636] | 531 | if ( ! ($token_modifiers[$i]&QST_NOT) ) |
---|
| 532 | continue; |
---|
| 533 | |
---|
| 534 | array_multisort($token_tag_scores[$i], SORT_DESC|SORT_NUMERIC, $token_tag_ids[$i]); |
---|
| 535 | |
---|
| 536 | for ($j=0; $j<count($token_tag_scores[$i]); $j++) |
---|
[10340] | 537 | { |
---|
[18636] | 538 | if ($token_tag_scores[$i][$j] < 0.8) |
---|
| 539 | break; |
---|
| 540 | if ($j>0 && $token_tag_scores[$i][$j] < $token_tag_scores[$i][0]) |
---|
| 541 | break; |
---|
| 542 | $tag_id = $token_tag_ids[$i][$j]; |
---|
| 543 | if ( isset($all_tags[$tag_id]) ) |
---|
| 544 | { |
---|
| 545 | unset($all_tags[$tag_id]); |
---|
| 546 | $not_tag_ids[] = $tag_id; |
---|
| 547 | } |
---|
| 548 | } |
---|
| 549 | $token_tag_ids[$i] = array(); |
---|
| 550 | } |
---|
| 551 | |
---|
| 552 | // process regular tags |
---|
| 553 | for ($i=0; $i<count($tokens); $i++) |
---|
| 554 | { |
---|
| 555 | if ( $token_modifiers[$i]&QST_NOT ) |
---|
| 556 | continue; |
---|
| 557 | |
---|
| 558 | array_multisort($token_tag_scores[$i], SORT_DESC|SORT_NUMERIC, $token_tag_ids[$i]); |
---|
| 559 | |
---|
| 560 | $counter = 0; |
---|
| 561 | for ($j=0; $j<count($token_tag_scores[$i]); $j++) |
---|
| 562 | { |
---|
| 563 | $tag_id = $token_tag_ids[$i][$j]; |
---|
| 564 | if ( ! isset($all_tags[$tag_id]) ) |
---|
| 565 | { |
---|
| 566 | array_splice($token_tag_ids[$i], $j, 1); |
---|
| 567 | array_splice($token_tag_scores[$i], $j, 1); |
---|
[22175] | 568 | $j--; |
---|
| 569 | continue; |
---|
[18636] | 570 | } |
---|
| 571 | |
---|
| 572 | $counter += $all_tags[$tag_id]['counter']; |
---|
| 573 | if ($counter > 200 && $j>0 && $token_tag_scores[$i][0] > $token_tag_scores[$i][$j] ) |
---|
[10340] | 574 | {// "many" images in previous tags and starting from this tag is less relevent |
---|
[18636] | 575 | array_splice($token_tag_ids[$i], $j); |
---|
| 576 | array_splice($token_tag_scores[$i], $j); |
---|
[10340] | 577 | break; |
---|
| 578 | } |
---|
| 579 | } |
---|
| 580 | } |
---|
[22175] | 581 | |
---|
[18636] | 582 | usort($all_tags, 'tag_alpha_compare'); |
---|
| 583 | foreach ( $all_tags as &$tag ) |
---|
[25658] | 584 | { |
---|
[18636] | 585 | $tag['name'] = trigger_event('render_tag_name', $tag['name']); |
---|
[25658] | 586 | } |
---|
[18636] | 587 | } |
---|
[10340] | 588 | |
---|
[18636] | 589 | /** |
---|
[25658] | 590 | * Returns the search results corresponding to a quick/query search. |
---|
[18636] | 591 | * A quick/query search returns many items (search is not strict), but results |
---|
| 592 | * are sorted by relevance unless $super_order_by is true. Returns: |
---|
[25658] | 593 | * array ( |
---|
| 594 | * 'items' => array of matching images |
---|
| 595 | * 'qs' => array( |
---|
| 596 | * 'matching_tags' => array of matching tags |
---|
| 597 | * 'matching_cats' => array of matching categories |
---|
| 598 | * 'matching_cats_no_images' =>array(99) - matching categories without images |
---|
| 599 | * ) |
---|
| 600 | * ) |
---|
[18636] | 601 | * |
---|
[25658] | 602 | * @param string $q |
---|
| 603 | * @param bool $super_order_by |
---|
| 604 | * @param string $images_where optional additional restriction on images table |
---|
[18636] | 605 | * @return array |
---|
| 606 | */ |
---|
| 607 | function get_quick_search_results($q, $super_order_by, $images_where='') |
---|
| 608 | { |
---|
| 609 | global $user, $conf; |
---|
| 610 | |
---|
| 611 | $search_results = |
---|
| 612 | array( |
---|
| 613 | 'items' => array(), |
---|
| 614 | 'qs' => array('q'=>stripslashes($q)), |
---|
| 615 | ); |
---|
| 616 | $q = trim($q); |
---|
| 617 | analyse_qsearch($q, $tokens, $token_modifiers); |
---|
| 618 | if (count($tokens)==0) |
---|
[10340] | 619 | { |
---|
[18636] | 620 | return $search_results; |
---|
| 621 | } |
---|
| 622 | $debug[] = '<!--'.count($tokens).' tokens'; |
---|
| 623 | |
---|
| 624 | $q_like_field = '@@__db_field__@@'; //something never in a search |
---|
| 625 | $q_like_clause = get_qsearch_like_clause($tokens, $token_modifiers, $q_like_field ); |
---|
| 626 | |
---|
| 627 | // Step 1 - first we find matches in #images table =========================== |
---|
| 628 | $where_clauses='MATCH(i.name, i.comment) AGAINST( \''.$q.'\' IN BOOLEAN MODE)'; |
---|
| 629 | if (!empty($q_like_clause)) |
---|
| 630 | { |
---|
| 631 | $where_clauses .= ' |
---|
| 632 | OR '. str_replace($q_like_field, 'CONVERT(file, CHAR)', $q_like_clause); |
---|
| 633 | $where_clauses = '('.$where_clauses.')'; |
---|
| 634 | } |
---|
| 635 | $where_clauses = array($where_clauses); |
---|
| 636 | if (!empty($images_where)) |
---|
| 637 | { |
---|
| 638 | $where_clauses[]='('.$images_where.')'; |
---|
| 639 | } |
---|
| 640 | $where_clauses[] .= get_sql_condition_FandF |
---|
| 641 | ( |
---|
| 642 | array( 'visible_images' => 'i.id' ), null, true |
---|
| 643 | ); |
---|
| 644 | $query = ' |
---|
| 645 | SELECT i.id, |
---|
| 646 | MATCH(i.name, i.comment) AGAINST( \''.$q.'\' IN BOOLEAN MODE) AS weight |
---|
| 647 | FROM '.IMAGES_TABLE.' i |
---|
| 648 | WHERE '.implode("\n AND ", $where_clauses); |
---|
| 649 | |
---|
| 650 | $by_weights=array(); |
---|
| 651 | $result = pwg_query($query); |
---|
| 652 | while ($row = pwg_db_fetch_assoc($result)) |
---|
| 653 | { // weight is important when sorting images by relevance |
---|
| 654 | if ($row['weight']) |
---|
| 655 | { |
---|
| 656 | $by_weights[(int)$row['id']] = 2*$row['weight']; |
---|
| 657 | } |
---|
| 658 | else |
---|
| 659 | {//full text does not match but file name match |
---|
| 660 | $by_weights[(int)$row['id']] = 2; |
---|
| 661 | } |
---|
| 662 | } |
---|
| 663 | $debug[] = count($by_weights).' fulltext'; |
---|
| 664 | if (!empty($by_weights)) |
---|
| 665 | { |
---|
| 666 | $debug[] = 'ft score min:'.min($by_weights).' max:'.max($by_weights); |
---|
| 667 | } |
---|
| 668 | |
---|
| 669 | |
---|
| 670 | // Step 2 - get the tags and the images for tags |
---|
| 671 | get_qsearch_tags($tokens, $token_modifiers, $token_tag_ids, $not_tag_ids, $search_results['qs']['matching_tags']); |
---|
| 672 | $debug[] = count($search_results['qs']['matching_tags']).' tags'; |
---|
| 673 | |
---|
| 674 | for ($i=0; $i<count($token_tag_ids); $i++) |
---|
| 675 | { |
---|
| 676 | $tag_ids = $token_tag_ids[$i]; |
---|
[18207] | 677 | $debug[] = count($tag_ids).' unique tags'; |
---|
[10340] | 678 | |
---|
| 679 | if (!empty($tag_ids)) |
---|
| 680 | { |
---|
[18207] | 681 | $tag_photo_count=0; |
---|
[1837] | 682 | $query = ' |
---|
[18636] | 683 | SELECT image_id FROM '.IMAGE_TAG_TABLE.' |
---|
[10340] | 684 | WHERE tag_id IN ('.implode(',',$tag_ids).') |
---|
[1837] | 685 | GROUP BY image_id'; |
---|
| 686 | $result = pwg_query($query); |
---|
[4325] | 687 | while ($row = pwg_db_fetch_assoc($result)) |
---|
[1837] | 688 | { // weight is important when sorting images by relevance |
---|
| 689 | $image_id=(int)$row['image_id']; |
---|
[10340] | 690 | @$by_weights[$image_id] += 1; |
---|
[18207] | 691 | $tag_photo_count++; |
---|
[1837] | 692 | } |
---|
[18636] | 693 | $debug[] = $tag_photo_count.' photos for tag'; |
---|
| 694 | $debug[] = count($by_weights).' photos after'; |
---|
[1837] | 695 | } |
---|
[1537] | 696 | } |
---|
| 697 | |
---|
[2135] | 698 | // Step 3 - search categories corresponding to the query $q ================== |
---|
| 699 | $query = ' |
---|
[2138] | 700 | SELECT id, name, permalink, nb_images |
---|
[2135] | 701 | FROM '.CATEGORIES_TABLE.' |
---|
| 702 | INNER JOIN '.USER_CACHE_CATEGORIES_TABLE.' ON id=cat_id |
---|
| 703 | WHERE user_id='.$user['id'].' |
---|
[6664] | 704 | AND MATCH(name, comment) AGAINST( \''.$q.'\' IN BOOLEAN MODE)'. |
---|
[2135] | 705 | get_sql_condition_FandF ( |
---|
| 706 | array( 'visible_categories' => 'cat_id' ), "\n AND" |
---|
| 707 | ); |
---|
| 708 | $result = pwg_query($query); |
---|
[4325] | 709 | while ($row = pwg_db_fetch_assoc($result)) |
---|
[2135] | 710 | { // weight is important when sorting images by relevance |
---|
| 711 | if ($row['nb_images']==0) |
---|
| 712 | { |
---|
[2138] | 713 | $search_results['qs']['matching_cats_no_images'][] = $row; |
---|
[2135] | 714 | } |
---|
| 715 | else |
---|
| 716 | { |
---|
[2138] | 717 | $search_results['qs']['matching_cats'][$row['id']] = $row; |
---|
[2135] | 718 | } |
---|
| 719 | } |
---|
[18207] | 720 | $debug[] = count(@$search_results['qs']['matching_cats']).' albums with images'; |
---|
[2135] | 721 | |
---|
| 722 | if ( empty($by_weights) and empty($search_results['qs']['matching_cats']) ) |
---|
[1537] | 723 | { |
---|
[2135] | 724 | return $search_results; |
---|
| 725 | } |
---|
| 726 | |
---|
[18636] | 727 | if (!empty($not_tag_ids)) |
---|
| 728 | { |
---|
| 729 | $query = ' |
---|
| 730 | SELECT image_id FROM '.IMAGE_TAG_TABLE.' |
---|
| 731 | WHERE tag_id IN ('.implode(',',$not_tag_ids).') |
---|
| 732 | GROUP BY image_id'; |
---|
| 733 | $result = pwg_query($query); |
---|
| 734 | while ($row = pwg_db_fetch_row($result)) |
---|
| 735 | { |
---|
| 736 | $id = $row[0]; |
---|
| 737 | unset($by_weights[$id]); |
---|
| 738 | } |
---|
| 739 | $debug[] = count($by_weights).' after not tags'; |
---|
| 740 | } |
---|
[2135] | 741 | // Step 4 - now we have $by_weights ( array image id => weight ) that need |
---|
| 742 | // permission checks and/or matching categories to get images from |
---|
| 743 | $where_clauses = array(); |
---|
| 744 | if ( !empty($by_weights) ) |
---|
| 745 | { |
---|
| 746 | $where_clauses[]='i.id IN (' |
---|
| 747 | . implode(',', array_keys($by_weights)) . ')'; |
---|
| 748 | } |
---|
| 749 | if ( !empty($search_results['qs']['matching_cats']) ) |
---|
| 750 | { |
---|
| 751 | $where_clauses[]='category_id IN ('. |
---|
[2138] | 752 | implode(',',array_keys($search_results['qs']['matching_cats'])).')'; |
---|
[2135] | 753 | } |
---|
| 754 | $where_clauses = array( '('.implode("\n OR ",$where_clauses).')' ); |
---|
| 755 | if (!empty($images_where)) |
---|
| 756 | { |
---|
| 757 | $where_clauses[]='('.$images_where.')'; |
---|
| 758 | } |
---|
| 759 | $where_clauses[] = get_sql_condition_FandF( |
---|
| 760 | array |
---|
| 761 | ( |
---|
| 762 | 'forbidden_categories' => 'category_id', |
---|
| 763 | 'visible_categories' => 'category_id', |
---|
| 764 | 'visible_images' => 'i.id' |
---|
| 765 | ), |
---|
| 766 | null,true |
---|
| 767 | ); |
---|
| 768 | |
---|
| 769 | $query = ' |
---|
[1537] | 770 | SELECT DISTINCT(id) |
---|
[2135] | 771 | FROM '.IMAGES_TABLE.' i |
---|
[1537] | 772 | INNER JOIN '.IMAGE_CATEGORY_TABLE.' AS ic ON id = ic.image_id |
---|
[2135] | 773 | WHERE '.implode("\n AND ", $where_clauses)."\n". |
---|
| 774 | $conf['order_by']; |
---|
| 775 | |
---|
| 776 | $allowed_images = array_from_query( $query, 'id'); |
---|
| 777 | |
---|
[18207] | 778 | $debug[] = count($allowed_images).' final photo count -->'; |
---|
| 779 | global $template; |
---|
| 780 | $template->append('footer_elements', implode(', ', $debug) ); |
---|
| 781 | |
---|
[2451] | 782 | if ( $super_order_by or empty($by_weights) ) |
---|
[2135] | 783 | { |
---|
| 784 | $search_results['items'] = $allowed_images; |
---|
| 785 | return $search_results; |
---|
[1537] | 786 | } |
---|
[2135] | 787 | |
---|
| 788 | $allowed_images = array_flip( $allowed_images ); |
---|
| 789 | $divisor = 5.0 * count($allowed_images); |
---|
[22175] | 790 | foreach ($allowed_images as $id=> &$rank ) |
---|
[1837] | 791 | { |
---|
[2135] | 792 | $weight = isset($by_weights[$id]) ? $by_weights[$id] : 1; |
---|
| 793 | $weight -= $rank/$divisor; |
---|
[22175] | 794 | $rank = $weight; |
---|
[1837] | 795 | } |
---|
[22175] | 796 | unset($rank); |
---|
| 797 | |
---|
[2135] | 798 | arsort($allowed_images, SORT_NUMERIC); |
---|
| 799 | $search_results['items'] = array_keys($allowed_images); |
---|
[1537] | 800 | return $search_results; |
---|
| 801 | } |
---|
| 802 | |
---|
| 803 | /** |
---|
[25658] | 804 | * Returns an array of 'items' corresponding to the search id. |
---|
| 805 | * It can be either a quick search or a regular search. |
---|
[1537] | 806 | * |
---|
[25658] | 807 | * @param int $search_id |
---|
| 808 | * @param bool $super_order_by |
---|
| 809 | * @param string $images_where optional aditional restriction on images table |
---|
[1537] | 810 | * @return array |
---|
| 811 | */ |
---|
[2451] | 812 | function get_search_results($search_id, $super_order_by, $images_where='') |
---|
[1537] | 813 | { |
---|
| 814 | $search = get_search_array($search_id); |
---|
| 815 | if ( !isset($search['q']) ) |
---|
| 816 | { |
---|
[2451] | 817 | $result['items'] = get_regular_search_results($search, $images_where); |
---|
[1537] | 818 | return $result; |
---|
| 819 | } |
---|
| 820 | else |
---|
| 821 | { |
---|
[2451] | 822 | return get_quick_search_results($search['q'], $super_order_by, $images_where); |
---|
[1537] | 823 | } |
---|
| 824 | } |
---|
[25658] | 825 | |
---|
[1113] | 826 | ?> |
---|