| 1 | <?php |
| 2 | /* -*- tab-width: 4; indent-tabs-mode: nil; c-basic-offset: 4 -*- */ |
| 3 | /* |
| 4 | # ***** BEGIN LICENSE BLOCK ***** |
| 5 | # This file is part of InDefero, an open source project management application. |
| 6 | # Copyright (C) 2008 Céondo Ltd and contributors. |
| 7 | # |
| 8 | # InDefero is free software; you can redistribute it and/or modify |
| 9 | # it under the terms of the GNU General Public License as published by |
| 10 | # the Free Software Foundation; either version 2 of the License, or |
| 11 | # (at your option) any later version. |
| 12 | # |
| 13 | # InDefero is distributed in the hope that it will be useful, |
| 14 | # but WITHOUT ANY WARRANTY; without even the implied warranty of |
| 15 | # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the |
| 16 | # GNU 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., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA |
| 21 | # |
| 22 | # Based on work under GNU LGPL copyright, from the Pluf Framework |
| 23 | # Copyright (C) 2001-2007 Loic d'Anterroches and contributors. |
| 24 | # |
| 25 | # ***** END LICENSE BLOCK ***** */ |
| 26 | |
| 27 | /** |
| 28 | * Class implementing the search engine |
| 29 | * |
| 30 | * It is a modified version of the Pluf_Search class to be able to |
| 31 | * cluster the results by project. |
| 32 | */ |
| 33 | class IDF_Search extends Pluf_Search |
| 34 | { |
| 35 | /** |
| 36 | * Search. |
| 37 | * |
| 38 | * Returns an array of array with model_class, model_id and |
| 39 | * score. The list is already sorted by score descending. |
| 40 | * |
| 41 | * You can then filter the list as you wish with another set of |
| 42 | * weights. |
| 43 | * |
| 44 | * @param string Query string. |
| 45 | * @param int Project id to limit the results (null) |
| 46 | * @param string Model class (null) |
| 47 | * @param string Stemmer class ('Pluf_Text_Stemmer_Porter') |
| 48 | * @return array Results |
| 49 | */ |
| 50 | public static function mySearch($query, $project=null, $model=null, $stemmer='Pluf_Text_Stemmer_Porter') |
| 51 | { |
| 52 | $query = Pluf_Text::cleanString(html_entity_decode($query, ENT_QUOTES, 'UTF-8')); |
| 53 | $words = Pluf_Text::tokenize($query); |
| 54 | if ($stemmer != null) { |
| 55 | $words = self::stem($words, $stemmer); |
| 56 | } |
| 57 | $words_flat = array(); |
| 58 | foreach ($words as $word=>$c) { |
| 59 | $words_flat[] = $word; |
| 60 | } |
| 61 | $word_ids = self::getWordIds($words_flat); |
| 62 | if (in_array(null, $word_ids) or count($word_ids) == 0) { |
| 63 | return array(); |
| 64 | } |
| 65 | return self::mySearchDocuments($word_ids, $project, $model); |
| 66 | } |
| 67 | |
| 68 | /** |
| 69 | * Search documents. |
| 70 | * |
| 71 | * Only the total of the ponderated occurences is used to sort the |
| 72 | * results. |
| 73 | * |
| 74 | * @param array Ids. |
| 75 | * @param IDF_Project Project to limit the search. |
| 76 | * @param string Model class to limit the search. |
| 77 | * @return array Sorted by score, returns model_class, model_id and score. |
| 78 | */ |
| 79 | public static function mySearchDocuments($wids, $project, $model) |
| 80 | { |
| 81 | $db =& Pluf::db(); |
| 82 | $gocc = new IDF_Search_Occ(); |
| 83 | $where = array(); |
| 84 | foreach ($wids as $id) { |
| 85 | $where[] = $db->qn('word').'='.(int)$id; |
| 86 | } |
| 87 | $prj = (is_null($project)) ? '' : ' AND project='.(int)$project->id; |
| 88 | $md = (is_null($model)) ? '' : ' AND model_class='.$db->esc($model); |
| 89 | $select = 'SELECT model_class, model_id, SUM(pondocc) AS score FROM '.$gocc->getSqlTable().' WHERE '.implode(' OR ', $where).$prj.$md.' GROUP BY model_class, model_id HAVING COUNT(*)='.count($wids).' ORDER BY score DESC'; |
| 90 | return $db->select($select); |
| 91 | } |
| 92 | |
| 93 | /** |
| 94 | * Index a document. |
| 95 | * |
| 96 | * See Pluf_Search for the disclaimer and informations. |
| 97 | * |
| 98 | * @param Pluf_Model Document to index. |
| 99 | * @param Stemmer used. ('Pluf_Text_Stemmer_Porter') |
| 100 | * @return array Statistics. |
| 101 | */ |
| 102 | public static function index($doc, $stemmer='Pluf_Text_Stemmer_Porter') |
| 103 | { |
| 104 | $words = Pluf_Text::tokenize($doc->_toIndex()); |
| 105 | if ($stemmer != null) { |
| 106 | $words = self::stem($words, $stemmer); |
| 107 | } |
| 108 | // Get the total number of words. |
| 109 | $total = 0.0; |
| 110 | $words_flat = array(); |
| 111 | foreach ($words as $word => $occ) { |
| 112 | $total += (float) $occ; |
| 113 | $words_flat[] = $word; |
| 114 | } |
| 115 | // Drop the last indexation. |
| 116 | $gocc = new IDF_Search_Occ(); |
| 117 | $sql = new Pluf_SQL('DELETE FROM '.$gocc->getSqlTable().' WHERE model_class=%s AND model_id=%s', array($doc->_model, $doc->id)); |
| 118 | $db =& Pluf::db(); |
| 119 | $db->execute($sql->gen()); |
| 120 | // Get the ids for each word. |
| 121 | $ids = self::getWordIds($words_flat); |
| 122 | // Insert a new word for the missing words and add the occ. |
| 123 | $n = count($ids); |
| 124 | $new_words = 0; |
| 125 | $done = array(); |
| 126 | for ($i=0;$i<$n;$i++) { |
| 127 | if ($ids[$i] === null) { |
| 128 | $word = new Pluf_Search_Word(); |
| 129 | $word->word = $words_flat[$i]; |
| 130 | try { |
| 131 | $word->create(); |
| 132 | $new_words++; |
| 133 | $ids[$i] = $word->id; |
| 134 | } catch (Exception $e) { |
| 135 | // 100% of the time, the word has been created |
| 136 | // by another process in the background. |
| 137 | $r_ids = self::getWordIds(array($word->word)); |
| 138 | if ($r_ids[0]) { |
| 139 | $ids[$i] = $r_ids[0]; |
| 140 | } else { |
| 141 | // give up for this word |
| 142 | continue; |
| 143 | } |
| 144 | } |
| 145 | } |
| 146 | if (isset($done[$ids[$i]])) { |
| 147 | continue; |
| 148 | } |
| 149 | $done[$ids[$i]] = true; |
| 150 | $occ = new IDF_Search_Occ(); |
| 151 | $occ->word = new Pluf_Search_Word($ids[$i]); |
| 152 | $occ->model_class = $doc->_model; |
| 153 | $occ->model_id = $doc->id; |
| 154 | $occ->project = $doc->get_project(); |
| 155 | $occ->occ = $words[$words_flat[$i]]; |
| 156 | $occ->pondocc = $words[$words_flat[$i]]/$total; |
| 157 | $occ->create(); |
| 158 | } |
| 159 | // update the stats |
| 160 | $sql = new Pluf_SQL('model_class=%s AND model_id=%s', |
| 161 | array($doc->_model, $doc->id)); |
| 162 | $last_index = Pluf::factory('Pluf_Search_Stats')->getList(array('filter' => $sql->gen())); |
| 163 | if ($last_index->count() == 0) { |
| 164 | $stats = new Pluf_Search_Stats(); |
| 165 | $stats->model_class = $doc->_model; |
| 166 | $stats->model_id = $doc->id; |
| 167 | $stats->indexations = 1; |
| 168 | $stats->create(); |
| 169 | } else { |
| 170 | $last_index[0]->indexations += 1; |
| 171 | $last_index[0]->update(); |
| 172 | } |
| 173 | return array('total' => $total, 'new' => $new_words, 'unique'=>$n); |
| 174 | } |
| 175 | |
| 176 | /** |
| 177 | * Remove an item from the index. |
| 178 | * |
| 179 | * You must call this function when you delete items wich are |
| 180 | * indexed. Just add the call: |
| 181 | * |
| 182 | * IDF_Search::remove($this); |
| 183 | * |
| 184 | * in the preDelete() method of your object. |
| 185 | * |
| 186 | * @param mixed Item to be removed |
| 187 | * @return bool Success |
| 188 | */ |
| 189 | public static function remove($item) |
| 190 | { |
| 191 | if ($item->id > 0) { |
| 192 | $sql = new Pluf_SQL('model_id=%s AND model_class=%s', |
| 193 | array($item->id, $item->_model)); |
| 194 | $items = Pluf::factory('IDF_Search_Occ')->getList(array('filter'=>$sql->gen())); |
| 195 | foreach ($items as $tl) { |
| 196 | $tl->delete(); |
| 197 | } |
| 198 | } |
| 199 | return true; |
| 200 | } |
| 201 | |
| 202 | } |
