includes/search/SearchOracle.php

   1 <?php
   2 # Copyright (C) 2004 Brion Vibber <brion@pobox.com>
   3 # http://www.mediawiki.org/
   4 #
   5 # This program is free software; you can redistribute it and/or modify
   6 # it under the terms of the GNU General Public License as published by
   7 # the Free Software Foundation; either version 2 of the License, or
   8 # (at your option) any later version.
   9 #
  10 # This program is distributed in the hope that it will be useful,
  11 # but WITHOUT ANY WARRANTY; without even the implied warranty of
  12 # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  13 # GNU General Public License for more details.
  14 #
  15 # You should have received a copy of the GNU General Public License along
  16 # with this program; if not, write to the Free Software Foundation, Inc.,
  17 # 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
  18 # http://www.gnu.org/copyleft/gpl.html
  19
  20 /**
  21  * @file
  22  * @ingroup Search
  23  */
  24
  25 /**
  26  * Search engine hook base class for Oracle (ConText).
  27  * @ingroup Search
  28  */
  29 class SearchOracle extends SearchEngine {
  30
  31         private $reservedWords = array ('ABOUT' => 1,
  32                                                                         'ACCUM' => 1,
  33                                                                         'AND' => 1,
  34                                                                         'BT' => 1,
  35                                                                         'BTG' => 1,
  36                                                                         'BTI' => 1,
  37                                                                         'BTP' => 1,
  38                                                                         'FUZZY' => 1,
  39                                                                         'HASPATH' => 1,
  40                                                                         'INPATH' => 1,
  41                                                                         'MINUS' => 1,
  42                                                                         'NEAR' => 1,
  43                                                                         'NOT' => 1,
  44                                                                         'NT' => 1,
  45                                                                         'NTG' => 1,
  46                                                                         'NTI' => 1,
  47                                                                         'NTP' => 1,
  48                                                                         'OR' => 1,
  49                                                                         'PT' => 1,
  50                                                                         'RT' => 1,
  51                                                                         'SQE' => 1,
  52                                                                         'SYN' => 1,
  53                                                                         'TR' => 1,
  54                                                                         'TRSYN' => 1,
  55                                                                         'TT' => 1,
  56                                                                         'WITHIN' => 1);
  57
  58         function __construct($db) {
  59                 $this->db = $db;
  60         }
  61
  62         /**
  63          * Perform a full text search query and return a result set.
  64          *
  65          * @param $term String: raw search term
  66          * @return SqlSearchResultSet
  67          */
  68         function searchText( $term ) {
  69                 if ($term == '')
  70                         return new SqlSearchResultSet(false, '');
  71
  72                 $resultSet = $this->db->resultObject($this->db->query($this->getQuery($this->filter($term), true)));
  73                 return new SqlSearchResultSet($resultSet, $this->searchTerms);
  74         }
  75
  76         /**
  77          * Perform a title-only search query and return a result set.
  78          *
  79          * @param $term String: raw search term
  80          * @return SqlSearchResultSet
  81          */
  82         function searchTitle($term) {
  83                 if ($term == '')
  84                         return new SqlSearchResultSet(false, '');
  85
  86                 $resultSet = $this->db->resultObject($this->db->query($this->getQuery($this->filter($term), false)));
  87                 return new MySQLSearchResultSet($resultSet, $this->searchTerms);
  88         }
  89
  90
  91         /**
  92          * Return a partial WHERE clause to exclude redirects, if so set
  93          * @return String
  94          */
  95         function queryRedirect() {
  96                 if ($this->showRedirects) {
  97                         return '';
  98                 } else {
  99                         return 'AND page_is_redirect=0';
 100                 }
 101         }
 102
 103         /**
 104          * Return a partial WHERE clause to limit the search to the given namespaces
 105          * @return String
 106          */
 107         function queryNamespaces() {
 108                 if( is_null($this->namespaces) )
 109                         return '';
 110                 if ( !count( $this->namespaces ) ) {
 111                         $namespaces = '0';
 112                 } else {
 113                         $namespaces = $this->db->makeList( $this->namespaces );
 114                 }
 115                 return 'AND page_namespace IN (' . $namespaces . ')';
 116         }
 117
 118         /**
 119          * Return a LIMIT clause to limit results on the query.
 120          * @return String
 121          */
 122         function queryLimit($sql) {
 123                 return $this->db->limitResult($sql, $this->limit, $this->offset);
 124         }
 125
 126         /**
 127          * Does not do anything for generic search engine
 128          * subclasses may define this though
 129          * @return String
 130          */
 131         function queryRanking($filteredTerm, $fulltext) {
 132                 return ' ORDER BY score(1)';
 133         }
 134
 135         /**
 136          * Construct the full SQL query to do the search.
 137          * The guts shoulds be constructed in queryMain()
 138          * @param $filteredTerm String
 139          * @param $fulltext Boolean
 140          */
 141         function getQuery( $filteredTerm, $fulltext ) {
 142                 return $this->queryLimit($this->queryMain($filteredTerm, $fulltext) . ' ' .
 143                         $this->queryRedirect() . ' ' .
 144                         $this->queryNamespaces() . ' ' .
 145                         $this->queryRanking( $filteredTerm, $fulltext ) . ' ');
 146         }
 147
 148
 149         /**
 150          * Picks which field to index on, depending on what type of query.
 151          * @param $fulltext Boolean
 152          * @return String
 153          */
 154         function getIndexField($fulltext) {
 155                 return $fulltext ? 'si_text' : 'si_title';
 156         }
 157
 158         /**
 159          * Get the base part of the search query.
 160          *
 161          * @param $filteredTerm String
 162          * @param $fulltext Boolean
 163          * @return String
 164          */
 165         function queryMain( $filteredTerm, $fulltext ) {
 166                 $match = $this->parseQuery($filteredTerm, $fulltext);
 167                 $page        = $this->db->tableName('page');
 168                 $searchindex = $this->db->tableName('searchindex');
 169                 return 'SELECT page_id, page_namespace, page_title ' .
 170                         "FROM $page,$searchindex " .
 171                         'WHERE page_id=si_page AND ' . $match;
 172         }
 173
 174         /**
 175          * Parse a user input search string, and return an SQL fragment to be used
 176          * as part of a WHERE clause
 177          */
 178         function parseQuery($filteredText, $fulltext) {
 179                 global $wgContLang;
 180                 $lc = SearchEngine::legalSearchChars();
 181                 $this->searchTerms = array();
 182
 183                 # FIXME: This doesn't handle parenthetical expressions.
 184                 $m = array();
 185                 $searchon = '';
 186                 if (preg_match_all('/([-+<>~]?)(([' . $lc . ']+)(\*?)|"[^"]*")/',
 187                           $filteredText, $m, PREG_SET_ORDER)) {
 188                         foreach($m as $terms) {
 189                                 // Search terms in all variant forms, only
 190                                 // apply on wiki with LanguageConverter
 191                                 $temp_terms = $wgContLang->autoConvertToAllVariants( $terms[2] );
 192                                 if( is_array( $temp_terms )) {
 193                                         $temp_terms = array_unique( array_values( $temp_terms ));
 194                                         foreach( $temp_terms as $t ) {
 195                                                 $searchon .= ($terms[1] == '-' ? ' ~' : ' & ') . $this->escapeTerm( $t );
 196                                         }
 197                                 }
 198                                 else {
 199                                         $searchon .= ($terms[1] == '-' ? ' ~' : ' & ') . $this->escapeTerm( $terms[2] );
 200                                 }
 201                                 if (!empty($terms[3])) {
 202                                         $regexp = preg_quote( $terms[3], '/' );
 203                                         if ($terms[4])
 204                                                 $regexp .= "[0-9A-Za-z_]+";
 205                                 } else {
 206                                         $regexp = preg_quote(str_replace('"', '', $terms[2]), '/');
 207                                 }
 208                                 $this->searchTerms[] = $regexp;
 209                         }
 210                 }
 211
 212
 213                 $searchon = $this->db->addQuotes(ltrim($searchon, ' &'));
 214                 $field = $this->getIndexField($fulltext);
 215                 return " CONTAINS($field, $searchon, 1) > 0 ";
 216         }
 217
 218         private function escapeTerm($t) {
 219                 global $wgContLang;
 220                 $t = $wgContLang->normalizeForSearch($t);
 221                 $t = isset($this->reservedWords[strtoupper($t)]) ? '{'.$t.'}' : $t;
 222                 $t = preg_replace('/^"(.*)"$/', '($1)', $t);
 223                 $t = preg_replace('/([-&|])/', '\\\\$1', $t);
 224                 return $t;
 225         }
 226         /**
 227          * Create or update the search index record for the given page.
 228          * Title and text should be pre-processed.
 229          *
 230          * @param $id Integer
 231          * @param $title String
 232          * @param $text String
 233          */
 234         function update($id, $title, $text) {
 235                 $dbw = wfGetDB(DB_MASTER);
 236                 $dbw->replace('searchindex',
 237                         array('si_page'),
 238                         array(
 239                                 'si_page' => $id,
 240                                 'si_title' => $title,
 241                                 'si_text' => $text
 242                         ), 'SearchOracle::update' );
 243                 $dbw->query("CALL ctx_ddl.sync_index('si_text_idx')");
 244                 $dbw->query("CALL ctx_ddl.sync_index('si_title_idx')");
 245         }
 246
 247         /**
 248          * Update a search index record's title only.
 249          * Title should be pre-processed.
 250          *
 251          * @param int $id
 252          * @param string $title
 253          */
 254         function updateTitle($id, $title) {
 255                 $dbw = wfGetDB(DB_MASTER);
 256
 257                 $dbw->update('searchindex',
 258                         array('si_title' => $title),
 259                         array('si_page'  => $id),
 260                         'SearchOracle::updateTitle',
 261                         array());
 262         }
 263
 264
 265         public static function legalSearchChars() {
 266                 return "\"" . parent::legalSearchChars();
 267         }
 268 }