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         /** @todo document */
  38         function parseQuery( $filteredText, $fulltext ) {
  39                 global $wgContLang;
  40                 $lc = SearchEngine::legalSearchChars(); // Minus format chars
  41                 $searchon = '';
  42                 $this->searchTerms = array();
  43
  44                 # FIXME: This doesn't handle parenthetical expressions.
  45                 $m = array();
  46                 if( preg_match_all( '/([-+<>~]?)(([' . $lc . ']+)(\*?)|"[^"]*")/',
  47                           $filteredText, $m, PREG_SET_ORDER ) ) {
  48                         foreach( $m as $terms ) {
  49                                 if( $searchon !== '' ) $searchon .= ' ';
  50                                 if( $this->strictMatching && ($terms[1] == '') ) {
  51                                         $terms[1] = '+';
  52                                 }
  53                                 $searchon .= $terms[1] . $wgContLang->stripForSearch( $terms[2] );
  54                                 if( !empty( $terms[3] ) ) {
  55                                         // Match individual terms in result highlighting...
  56                                         $regexp = preg_quote( $terms[3], '/' );
  57                                         if( $terms[4] ) {
  58                                                 $regexp = "\b$regexp"; // foo*
  59                                         } else {
  60                                                 $regexp = "\b$regexp\b";
  61                                         }
  62                                 } else {
  63                                         // Match the quoted term in result highlighting...
  64                                         $regexp = preg_quote( str_replace( '"', '', $terms[2] ), '/' );
  65                                 }
  66                                 $this->searchTerms[] = $regexp;
  67                         }
  68                         wfDebug( "Would search with '$searchon'\n" );
  69                         wfDebug( 'Match with /' . implode( '|', $this->searchTerms ) . "/\n" );
  70                 } else {
  71                         wfDebug( "Can't understand search query '{$filteredText}'\n" );
  72                 }
  73
  74                 $searchon = $this->db->strencode( $searchon );
  75                 $field = $this->getIndexField( $fulltext );
  76                 return " MATCH($field) AGAINST('$searchon' IN BOOLEAN MODE) ";
  77         }
  78
  79         public static function legalSearchChars() {
  80                 return "\"*" . parent::legalSearchChars();
  81         }
  82
  83         /**
  84          * Perform a full text search query and return a result set.
  85          *
  86          * @param string $term - Raw search term
  87          * @return MySQLSearchResultSet
  88          * @access public
  89          */
  90         function searchText( $term ) {
  91                 $resultSet = $this->db->resultObject( $this->db->query( $this->getQuery( $this->filter( $term ), true ) ) );
  92                 return new MySQLSearchResultSet( $resultSet, $this->searchTerms );
  93         }
  94
  95         /**
  96          * Perform a title-only search query and return a result set.
  97          *
  98          * @param string $term - Raw search term
  99          * @return MySQLSearchResultSet
 100          * @access public
 101          */
 102         function searchTitle( $term ) {
 103                 $resultSet = $this->db->resultObject( $this->db->query( $this->getQuery( $this->filter( $term ), false ) ) );
 104                 return new MySQLSearchResultSet( $resultSet, $this->searchTerms );
 105         }
 106
 107
 108         /**
 109          * Return a partial WHERE clause to exclude redirects, if so set
 110          * @return string
 111          * @private
 112          */
 113         function queryRedirect() {
 114                 if( $this->showRedirects ) {
 115                         return '';
 116                 } else {
 117                         return 'AND page_is_redirect=0';
 118                 }
 119         }
 120
 121         /**
 122          * Return a partial WHERE clause to limit the search to the given namespaces
 123          * @return string
 124          * @private
 125          */
 126         function queryNamespaces() {
 127                 if( is_null($this->namespaces) )
 128                         return '';  # search all
 129                 $namespaces = implode( ',', $this->namespaces );
 130                 if ($namespaces == '') {
 131                         $namespaces = '0';
 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          * @private
 140          */
 141         function queryLimit() {
 142                 return $this->db->limitResult( '', $this->limit, $this->offset );
 143         }
 144
 145         /**
 146          * Does not do anything for generic search engine
 147          * subclasses may define this though
 148          * @return string
 149          * @private
 150          */
 151         function queryRanking( $filteredTerm, $fulltext ) {
 152                 return '';
 153         }
 154
 155         /**
 156          * Construct the full SQL query to do the search.
 157          * The guts shoulds be constructed in queryMain()
 158          * @param string $filteredTerm
 159          * @param bool $fulltext
 160          * @private
 161          */
 162         function getQuery( $filteredTerm, $fulltext ) {
 163                 return $this->queryMain( $filteredTerm, $fulltext ) . ' ' .
 164                         $this->queryRedirect() . ' ' .
 165                         $this->queryNamespaces() . ' ' .
 166                         $this->queryRanking( $filteredTerm, $fulltext ) . ' ' .
 167                         $this->queryLimit();
 168         }
 169
 170
 171         /**
 172          * Picks which field to index on, depending on what type of query.
 173          * @param bool $fulltext
 174          * @return string
 175          */
 176         function getIndexField( $fulltext ) {
 177                 return $fulltext ? 'si_text' : 'si_title';
 178         }
 179
 180         /**
 181          * Get the base part of the search query.
 182          * The actual match syntax will depend on the server
 183          * version; MySQL 3 and MySQL 4 have different capabilities
 184          * in their fulltext search indexes.
 185          *
 186          * @param string $filteredTerm
 187          * @param bool $fulltext
 188          * @return string
 189          * @private
 190          */
 191         function queryMain( $filteredTerm, $fulltext ) {
 192                 $match = $this->parseQuery( $filteredTerm, $fulltext );
 193                 $page        = $this->db->tableName( 'page' );
 194                 $searchindex = $this->db->tableName( 'searchindex' );
 195                 return 'SELECT page_id, page_namespace, page_title ' .
 196                         "FROM $page,$searchindex " .
 197                         'WHERE page_id=si_page AND ' . $match;
 198         }
 199
 200         /**
 201          * Create or update the search index record for the given page.
 202          * Title and text should be pre-processed.
 203          *
 204          * @param int $id
 205          * @param string $title
 206          * @param string $text
 207          */
 208         function update( $id, $title, $text ) {
 209                 $dbw = wfGetDB( DB_MASTER );
 210                 $dbw->replace( 'searchindex',
 211                         array( 'si_page' ),
 212                         array(
 213                                 'si_page' => $id,
 214                                 'si_title' => $title,
 215                                 'si_text' => $text
 216                         ), __METHOD__ );
 217         }
 218
 219         /**
 220          * Update a search index record's title only.
 221          * Title should be pre-processed.
 222          *
 223          * @param int $id
 224          * @param string $title
 225          */
 226     function updateTitle( $id, $title ) {
 227                 $dbw = wfGetDB( DB_MASTER );
 228
 229                 $dbw->update( 'searchindex',
 230                         array( 'si_title' => $title ),
 231                         array( 'si_page'  => $id ),
 232                         __METHOD__,
 233                         array( $dbw->lowPriorityOption() ) );
 234         }
 235 }
 236
 237 /**
 238  * @ingroup Search
 239  */
 240 class MySQLSearchResultSet extends SearchResultSet {
 241         function MySQLSearchResultSet( $resultSet, $terms ) {
 242                 $this->mResultSet = $resultSet;
 243                 $this->mTerms = $terms;
 244         }
 245
 246         function termMatches() {
 247                 return $this->mTerms;
 248         }
 249
 250         function numRows() {
 251                 return $this->mResultSet->numRows();
 252         }
 253
 254         function next() {
 255                 $row = $this->mResultSet->fetchObject();
 256                 if( $row === false ) {
 257                         return false;
 258                 } else {
 259                         return new SearchResult( $row );
 260                 }
 261         }
 262
 263         function free() {
 264                 $this->mResultSet->free();
 265         }
 266 }