includes/SearchMySQL.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  * Search engine hook for MySQL 4+
  22  * @addtogroup Search
  23  */
  24 class SearchMySQL extends SearchEngine {
  25         var $strictMatching = true;
  26
  27         /** @todo document */
  28         function __construct( $db ) {
  29                 $this->db = $db;
  30         }
  31
  32         /** @todo document */
  33         function parseQuery( $filteredText, $fulltext ) {
  34                 global $wgContLang;
  35                 $lc = SearchEngine::legalSearchChars(); // Minus format chars
  36                 $searchon = '';
  37                 $this->searchTerms = array();
  38
  39                 # FIXME: This doesn't handle parenthetical expressions.
  40                 $m = array();
  41                 if( preg_match_all( '/([-+<>~]?)(([' . $lc . ']+)(\*?)|"[^"]*")/',
  42                           $filteredText, $m, PREG_SET_ORDER ) ) {
  43                         foreach( $m as $terms ) {
  44                                 if( $searchon !== '' ) $searchon .= ' ';
  45                                 if( $this->strictMatching && ($terms[1] == '') ) {
  46                                         $terms[1] = '+';
  47                                 }
  48                                 $searchon .= $terms[1] . $wgContLang->stripForSearch( $terms[2] );
  49                                 if( !empty( $terms[3] ) ) {
  50                                         // Match individual terms in result highlighting...
  51                                         $regexp = preg_quote( $terms[3], '/' );
  52                                         if( $terms[4] ) $regexp .= "[0-9A-Za-z_]+";
  53                                 } else {
  54                                         // Match the quoted term in result highlighting...
  55                                         $regexp = preg_quote( str_replace( '"', '', $terms[2] ), '/' );
  56                                 }
  57                                 $this->searchTerms[] = "\b$regexp\b";
  58                         }
  59                         wfDebug( "Would search with '$searchon'\n" );
  60                         wfDebug( 'Match with /\b' . implode( '\b|\b', $this->searchTerms ) . "\b/\n" );
  61                 } else {
  62                         wfDebug( "Can't understand search query '{$filteredText}'\n" );
  63                 }
  64
  65                 $searchon = $this->db->strencode( $searchon );
  66                 $field = $this->getIndexField( $fulltext );
  67                 return " MATCH($field) AGAINST('$searchon' IN BOOLEAN MODE) ";
  68         }
  69
  70         public static function legalSearchChars() {
  71                 return "\"*" . parent::legalSearchChars();
  72         }
  73
  74         /**
  75          * Perform a full text search query and return a result set.
  76          *
  77          * @param string $term - Raw search term
  78          * @return MySQLSearchResultSet
  79          * @access public
  80          */
  81         function searchText( $term ) {
  82                 $resultSet = $this->db->resultObject( $this->db->query( $this->getQuery( $this->filter( $term ), true ) ) );
  83                 return new MySQLSearchResultSet( $resultSet, $this->searchTerms );
  84         }
  85
  86         /**
  87          * Perform a title-only search query and return a result set.
  88          *
  89          * @param string $term - Raw search term
  90          * @return MySQLSearchResultSet
  91          * @access public
  92          */
  93         function searchTitle( $term ) {
  94                 $resultSet = $this->db->resultObject( $this->db->query( $this->getQuery( $this->filter( $term ), false ) ) );
  95                 return new MySQLSearchResultSet( $resultSet, $this->searchTerms );
  96         }
  97
  98
  99         /**
 100          * Return a partial WHERE clause to exclude redirects, if so set
 101          * @return string
 102          * @private
 103          */
 104         function queryRedirect() {
 105                 if( $this->showRedirects ) {
 106                         return '';
 107                 } else {
 108                         return 'AND page_is_redirect=0';
 109                 }
 110         }
 111
 112         /**
 113          * Return a partial WHERE clause to limit the search to the given namespaces
 114          * @return string
 115          * @private
 116          */
 117         function queryNamespaces() {
 118                 $namespaces = implode( ',', $this->namespaces );
 119                 if ($namespaces == '') {
 120                         $namespaces = '0';
 121                 }
 122                 return 'AND page_namespace IN (' . $namespaces . ')';
 123         }
 124
 125         /**
 126          * Return a LIMIT clause to limit results on the query.
 127          * @return string
 128          * @private
 129          */
 130         function queryLimit() {
 131                 return $this->db->limitResult( '', $this->limit, $this->offset );
 132         }
 133
 134         /**
 135          * Does not do anything for generic search engine
 136          * subclasses may define this though
 137          * @return string
 138          * @private
 139          */
 140         function queryRanking( $filteredTerm, $fulltext ) {
 141                 return '';
 142         }
 143
 144         /**
 145          * Construct the full SQL query to do the search.
 146          * The guts shoulds be constructed in queryMain()
 147          * @param string $filteredTerm
 148          * @param bool $fulltext
 149          * @private
 150          */
 151         function getQuery( $filteredTerm, $fulltext ) {
 152                 return $this->queryMain( $filteredTerm, $fulltext ) . ' ' .
 153                         $this->queryRedirect() . ' ' .
 154                         $this->queryNamespaces() . ' ' .
 155                         $this->queryRanking( $filteredTerm, $fulltext ) . ' ' .
 156                         $this->queryLimit();
 157         }
 158
 159
 160         /**
 161          * Picks which field to index on, depending on what type of query.
 162          * @param bool $fulltext
 163          * @return string
 164          */
 165         function getIndexField( $fulltext ) {
 166                 return $fulltext ? 'si_text' : 'si_title';
 167         }
 168
 169         /**
 170          * Get the base part of the search query.
 171          * The actual match syntax will depend on the server
 172          * version; MySQL 3 and MySQL 4 have different capabilities
 173          * in their fulltext search indexes.
 174          *
 175          * @param string $filteredTerm
 176          * @param bool $fulltext
 177          * @return string
 178          * @private
 179          */
 180         function queryMain( $filteredTerm, $fulltext ) {
 181                 $match = $this->parseQuery( $filteredTerm, $fulltext );
 182                 $page        = $this->db->tableName( 'page' );
 183                 $searchindex = $this->db->tableName( 'searchindex' );
 184                 return 'SELECT page_id, page_namespace, page_title ' .
 185                         "FROM $page,$searchindex " .
 186                         'WHERE page_id=si_page AND ' . $match;
 187         }
 188
 189         /**
 190          * Create or update the search index record for the given page.
 191          * Title and text should be pre-processed.
 192          *
 193          * @param int $id
 194          * @param string $title
 195          * @param string $text
 196          */
 197         function update( $id, $title, $text ) {
 198                 $dbw = wfGetDB( DB_MASTER );
 199                 $dbw->replace( 'searchindex',
 200                         array( 'si_page' ),
 201                         array(
 202                                 'si_page' => $id,
 203                                 'si_title' => $title,
 204                                 'si_text' => $text
 205                         ), __METHOD__ );
 206         }
 207
 208         /**
 209          * Update a search index record's title only.
 210          * Title should be pre-processed.
 211          *
 212          * @param int $id
 213          * @param string $title
 214          */
 215     function updateTitle( $id, $title ) {
 216                 $dbw = wfGetDB( DB_MASTER );
 217
 218                 $dbw->update( 'searchindex',
 219                         array( 'si_title' => $title ),
 220                         array( 'si_page'  => $id ),
 221                         __METHOD__,
 222                         array( $dbw->lowPriorityOption() ) );
 223         }
 224 }
 225
 226 /**
 227  * @addtogroup Search
 228  */
 229 class MySQLSearchResultSet extends SearchResultSet {
 230         function MySQLSearchResultSet( $resultSet, $terms ) {
 231                 $this->mResultSet = $resultSet;
 232                 $this->mTerms = $terms;
 233         }
 234
 235         function termMatches() {
 236                 return $this->mTerms;
 237         }
 238
 239         function numRows() {
 240                 return $this->mResultSet->numRows();
 241         }
 242
 243         function next() {
 244                 $row = $this->mResultSet->fetchObject();
 245                 if( $row === false ) {
 246                         return false;
 247                 } else {
 248                         return new SearchResult( $row );
 249                 }
 250         }
 251
 252         function free() {
 253                 $this->mResultSet->free();
 254         }
 255 }
 256
 257