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                                 // Search terms in all variant forms, only
  57                                 // apply on wiki with LanguageConverter
  58                                 $temp_terms = $wgContLang->autoConvertToAllVariants( $terms[2] );
  59                                 if( is_array( $temp_terms )) {
  60                                         $temp_terms = array_unique( array_values( $temp_terms ));
  61                                         foreach( $temp_terms as $t )
  62                                                 $searchon .= $terms[1] . $wgContLang->stripForSearch( $t ) . ' ';
  63                                 }
  64                                 else
  65                                         $searchon .= $terms[1] . $wgContLang->stripForSearch( $terms[2] );
  66                                 if( !empty( $terms[3] ) ) {
  67                                         // Match individual terms in result highlighting...
  68                                         $regexp = preg_quote( $terms[3], '/' );
  69                                         if( $terms[4] ) {
  70                                                 $regexp = "\b$regexp"; // foo*
  71                                         } else {
  72                                                 $regexp = "\b$regexp\b";
  73                                         }
  74                                 } else {
  75                                         // Match the quoted term in result highlighting...
  76                                         $regexp = preg_quote( str_replace( '"', '', $terms[2] ), '/' );
  77                                 }
  78                                 $this->searchTerms[] = $regexp;
  79                         }
  80                         wfDebug( "Would search with '$searchon'\n" );
  81                         wfDebug( 'Match with /' . implode( '|', $this->searchTerms ) . "/\n" );
  82                 } else {
  83                         wfDebug( "Can't understand search query '{$filteredText}'\n" );
  84                 }
  85
  86                 $searchon = $this->db->strencode( $searchon );
  87                 $field = $this->getIndexField( $fulltext );
  88                 return " MATCH($field) AGAINST('$searchon' IN BOOLEAN MODE) ";
  89         }
  90
  91         public static function legalSearchChars() {
  92                 return "\"*" . parent::legalSearchChars();
  93         }
  94
  95         /**
  96          * Perform a full text search query and return a result set.
  97          *
  98          * @param $term String: raw search term
  99          * @return MySQLSearchResultSet
 100          */
 101         function searchText( $term ) {
 102                 $resultSet = $this->db->resultObject( $this->db->query( $this->getQuery( $this->filter( $term ), true ) ) );
 103                 return new MySQLSearchResultSet( $resultSet, $this->searchTerms );
 104         }
 105
 106         /**
 107          * Perform a title-only search query and return a result set.
 108          *
 109          * @param $term String: raw search term
 110          * @return MySQLSearchResultSet
 111          */
 112         function searchTitle( $term ) {
 113                 $resultSet = $this->db->resultObject( $this->db->query( $this->getQuery( $this->filter( $term ), false ) ) );
 114                 return new MySQLSearchResultSet( $resultSet, $this->searchTerms );
 115         }
 116
 117
 118         /**
 119          * Return a partial WHERE clause to exclude redirects, if so set
 120          * @return String
 121          */
 122         function queryRedirect() {
 123                 if( $this->showRedirects ) {
 124                         return '';
 125                 } else {
 126                         return 'AND page_is_redirect=0';
 127                 }
 128         }
 129
 130         /**
 131          * Return a partial WHERE clause to limit the search to the given namespaces
 132          * @return String
 133          */
 134         function queryNamespaces() {
 135                 if( is_null($this->namespaces) )
 136                         return '';  # search all
 137                 if ( !count( $this->namespaces ) ) {
 138                         $namespaces = '0';
 139                 } else {
 140                         $namespaces = $this->db->makeList( $this->namespaces );
 141                 }
 142                 return 'AND page_namespace IN (' . $namespaces . ')';
 143         }
 144
 145         /**
 146          * Return a LIMIT clause to limit results on the query.
 147          * @return String
 148          */
 149         function queryLimit() {
 150                 return $this->db->limitResult( '', $this->limit, $this->offset );
 151         }
 152
 153         /**
 154          * Does not do anything for generic search engine
 155          * subclasses may define this though
 156          * @return String
 157          */
 158         function queryRanking( $filteredTerm, $fulltext ) {
 159                 return '';
 160         }
 161
 162         /**
 163          * Construct the full SQL query to do the search.
 164          * The guts shoulds be constructed in queryMain()
 165          * @param $filteredTerm String
 166          * @param $fulltext Boolean
 167          */
 168         function getQuery( $filteredTerm, $fulltext ) {
 169                 return $this->queryMain( $filteredTerm, $fulltext ) . ' ' .
 170                         $this->queryRedirect() . ' ' .
 171                         $this->queryNamespaces() . ' ' .
 172                         $this->queryRanking( $filteredTerm, $fulltext ) . ' ' .
 173                         $this->queryLimit();
 174         }
 175
 176
 177         /**
 178          * Picks which field to index on, depending on what type of query.
 179          * @param $fulltext Boolean
 180          * @return String
 181          */
 182         function getIndexField( $fulltext ) {
 183                 return $fulltext ? 'si_text' : 'si_title';
 184         }
 185
 186         /**
 187          * Get the base part of the search query.
 188          * The actual match syntax will depend on the server
 189          * version; MySQL 3 and MySQL 4 have different capabilities
 190          * in their fulltext search indexes.
 191          *
 192          * @param $filteredTerm String
 193          * @param $fulltext Boolean
 194          * @return String
 195          */
 196         function queryMain( $filteredTerm, $fulltext ) {
 197                 $match = $this->parseQuery( $filteredTerm, $fulltext );
 198                 $page        = $this->db->tableName( 'page' );
 199                 $searchindex = $this->db->tableName( 'searchindex' );
 200                 return 'SELECT page_id, page_namespace, page_title ' .
 201                         "FROM $page,$searchindex " .
 202                         'WHERE page_id=si_page AND ' . $match;
 203         }
 204
 205         /**
 206          * Create or update the search index record for the given page.
 207          * Title and text should be pre-processed.
 208          *
 209          * @param $id Integer
 210          * @param $title String
 211          * @param $text String
 212          */
 213         function update( $id, $title, $text ) {
 214                 $dbw = wfGetDB( DB_MASTER );
 215                 $dbw->replace( 'searchindex',
 216                         array( 'si_page' ),
 217                         array(
 218                                 'si_page' => $id,
 219                                 'si_title' => $title,
 220                                 'si_text' => $text
 221                         ), __METHOD__ );
 222         }
 223
 224         /**
 225          * Update a search index record's title only.
 226          * Title should be pre-processed.
 227          *
 228          * @param $id Integer
 229          * @param $title String
 230          */
 231     function updateTitle( $id, $title ) {
 232                 $dbw = wfGetDB( DB_MASTER );
 233
 234                 $dbw->update( 'searchindex',
 235                         array( 'si_title' => $title ),
 236                         array( 'si_page'  => $id ),
 237                         __METHOD__,
 238                         array( $dbw->lowPriorityOption() ) );
 239         }
 240 }
 241
 242 /**
 243  * @ingroup Search
 244  */
 245 class MySQLSearchResultSet extends SearchResultSet {
 246         function MySQLSearchResultSet( $resultSet, $terms ) {
 247                 $this->mResultSet = $resultSet;
 248                 $this->mTerms = $terms;
 249         }
 250
 251         function termMatches() {
 252                 return $this->mTerms;
 253         }
 254
 255         function numRows() {
 256                 return $this->mResultSet->numRows();
 257         }
 258
 259         function next() {
 260                 $row = $this->mResultSet->fetchObject();
 261                 if( $row === false ) {
 262                         return false;
 263                 } else {
 264                         return new SearchResult( $row );
 265                 }
 266         }
 267
 268         function free() {
 269                 $this->mResultSet->free();
 270         }
 271 }