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  * @file
  22  * @ingroup Search
  23  */
  24
  25 /**
  26  * Search engine hook for MySQL 4+
  27  * @ingroup Search
  28  */
  29 class SearchMySQL extends SearchEngine {
  30         var $strictMatching = true;
  31
  32         /** @todo document */
  33         function __construct( $db ) {
  34                 $this->db = $db;
  35         }
  36
  37         /**
  38          * Parse the user's query and transform it into an SQL fragment which will
  39          * become part of a WHERE clause
  40          */
  41         function parseQuery( $filteredText, $fulltext ) {
  42                 global $wgContLang;
  43                 $lc = SearchEngine::legalSearchChars(); // Minus format chars
  44                 $searchon = '';
  45                 $this->searchTerms = array();
  46
  47                 # FIXME: This doesn't handle parenthetical expressions.
  48                 $m = array();
  49                 if( preg_match_all( '/([-+<>~]?)(([' . $lc . ']+)(\*?)|"[^"]*")/',
  50                           $filteredText, $m, PREG_SET_ORDER ) ) {
  51                         foreach( $m as $terms ) {
  52                                 if( $searchon !== '' ) $searchon .= ' ';
  53                                 if( $this->strictMatching && ($terms[1] == '') ) {
  54                                         $terms[1] = '+';
  55                                 }
  56                                 $searchon .= $terms[1] . $wgContLang->stripForSearch( $terms[2] );
  57                                 if( !empty( $terms[3] ) ) {
  58                                         // Match individual terms in result highlighting...
  59                                         $regexp = preg_quote( $terms[3], '/' );
  60                                         if( $terms[4] ) {
  61                                                 $regexp = "\b$regexp"; // foo*
  62                                         } else {
  63                                                 $regexp = "\b$regexp\b";
  64                                         }
  65                                 } else {
  66                                         // Match the quoted term in result highlighting...
  67                                         $regexp = preg_quote( str_replace( '"', '', $terms[2] ), '/' );
  68                                 }
  69                                 $this->searchTerms[] = $regexp;
  70                         }
  71                         wfDebug( "Would search with '$searchon'\n" );
  72                         wfDebug( 'Match with /' . implode( '|', $this->searchTerms ) . "/\n" );
  73                 } else {
  74                         wfDebug( "Can't understand search query '{$filteredText}'\n" );
  75                 }
  76
  77                 $searchon = $this->db->strencode( $searchon );
  78                 $field = $this->getIndexField( $fulltext );
  79                 return " MATCH($field) AGAINST('$searchon' IN BOOLEAN MODE) ";
  80         }
  81
  82         public static function legalSearchChars() {
  83                 return "\"*" . parent::legalSearchChars();
  84         }
  85
  86         /**
  87          * Perform a full text search query and return a result set.
  88          *
  89          * @param $term String: raw search term
  90          * @return MySQLSearchResultSet
  91          */
  92         function searchText( $term ) {
  93                 $resultSet = $this->db->resultObject( $this->db->query( $this->getQuery( $this->filter( $term ), true ) ) );
  94                 return new MySQLSearchResultSet( $resultSet, $this->searchTerms );
  95         }
  96
  97         /**
  98          * Perform a title-only search query and return a result set.
  99          *
 100          * @param $term String: raw search term
 101          * @return MySQLSearchResultSet
 102          */
 103         function searchTitle( $term ) {
 104                 $resultSet = $this->db->resultObject( $this->db->query( $this->getQuery( $this->filter( $term ), false ) ) );
 105                 return new MySQLSearchResultSet( $resultSet, $this->searchTerms );
 106         }
 107
 108
 109         /**
 110          * Return a partial WHERE clause to exclude redirects, if so set
 111          * @return String
 112          */
 113         function queryRedirect() {
 114                 if( $this->showRedirects ) {
 115                         return '';
 116                 } else {
 117                         return 'AND page_is_redirect=0';
 118                 }
 119         }
 120
 121         /**
 122          * Return a partial WHERE clause to limit the search to the given namespaces
 123          * @return String
 124          */
 125         function queryNamespaces() {
 126                 if( is_null($this->namespaces) )
 127                         return '';  # search all
 128                 if ( !count( $this->namespaces ) ) {
 129                         $namespaces = '0';
 130                 } else {
 131                         $namespaces = $this->db->makeList( $this->namespaces );
 132                 }
 133                 return 'AND page_namespace IN (' . $namespaces . ')';
 134         }
 135
 136         /**
 137          * Return a LIMIT clause to limit results on the query.
 138          * @return String
 139          */
 140         function queryLimit() {
 141                 return $this->db->limitResult( '', $this->limit, $this->offset );
 142         }
 143
 144         /**
 145          * Does not do anything for generic search engine
 146          * subclasses may define this though
 147          * @return String
 148          */
 149         function queryRanking( $filteredTerm, $fulltext ) {
 150                 return '';
 151         }
 152
 153         /**
 154          * Construct the full SQL query to do the search.
 155          * The guts shoulds be constructed in queryMain()
 156          * @param $filteredTerm String
 157          * @param $fulltext Boolean
 158          */
 159         function getQuery( $filteredTerm, $fulltext ) {
 160                 return $this->queryMain( $filteredTerm, $fulltext ) . ' ' .
 161                         $this->queryRedirect() . ' ' .
 162                         $this->queryNamespaces() . ' ' .
 163                         $this->queryRanking( $filteredTerm, $fulltext ) . ' ' .
 164                         $this->queryLimit();
 165         }
 166
 167
 168         /**
 169          * Picks which field to index on, depending on what type of query.
 170          * @param $fulltext Boolean
 171          * @return String
 172          */
 173         function getIndexField( $fulltext ) {
 174                 return $fulltext ? 'si_text' : 'si_title';
 175         }
 176
 177         /**
 178          * Get the base part of the search query.
 179          * The actual match syntax will depend on the server
 180          * version; MySQL 3 and MySQL 4 have different capabilities
 181          * in their fulltext search indexes.
 182          *
 183          * @param $filteredTerm String
 184          * @param $fulltext Boolean
 185          * @return String
 186          */
 187         function queryMain( $filteredTerm, $fulltext ) {
 188                 $match = $this->parseQuery( $filteredTerm, $fulltext );
 189                 $page        = $this->db->tableName( 'page' );
 190                 $searchindex = $this->db->tableName( 'searchindex' );
 191                 return 'SELECT page_id, page_namespace, page_title ' .
 192                         "FROM $page,$searchindex " .
 193                         'WHERE page_id=si_page AND ' . $match;
 194         }
 195
 196         /**
 197          * Create or update the search index record for the given page.
 198          * Title and text should be pre-processed.
 199          *
 200          * @param $id Integer
 201          * @param $title String
 202          * @param $text String
 203          */
 204         function update( $id, $title, $text ) {
 205                 $dbw = wfGetDB( DB_MASTER );
 206                 $dbw->replace( 'searchindex',
 207                         array( 'si_page' ),
 208                         array(
 209                                 'si_page' => $id,
 210                                 'si_title' => $title,
 211                                 'si_text' => $text
 212                         ), __METHOD__ );
 213         }
 214
 215         /**
 216          * Update a search index record's title only.
 217          * Title should be pre-processed.
 218          *
 219          * @param $id Integer
 220          * @param $title String
 221          */
 222     function updateTitle( $id, $title ) {
 223                 $dbw = wfGetDB( DB_MASTER );
 224
 225                 $dbw->update( 'searchindex',
 226                         array( 'si_title' => $title ),
 227                         array( 'si_page'  => $id ),
 228                         __METHOD__,
 229                         array( $dbw->lowPriorityOption() ) );
 230         }
 231 }
 232
 233 /**
 234  * @ingroup Search
 235  */
 236 class MySQLSearchResultSet extends SearchResultSet {
 237         function MySQLSearchResultSet( $resultSet, $terms ) {
 238                 $this->mResultSet = $resultSet;
 239                 $this->mTerms = $terms;
 240         }
 241
 242         function termMatches() {
 243                 return $this->mTerms;
 244         }
 245
 246         function numRows() {
 247                 return $this->mResultSet->numRows();
 248         }
 249
 250         function next() {
 251                 $row = $this->mResultSet->fetchObject();
 252                 if( $row === false ) {
 253                         return false;
 254                 } else {
 255                         return new SearchResult( $row );
 256                 }
 257         }
 258
 259         function free() {
 260                 $this->mResultSet->free();
 261         }
 262 }