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 string $term - Raw search term
  90          * @return MySQLSearchResultSet
  91          * @access public
  92          */
  93         function searchText( $term ) {
  94                 $resultSet = $this->db->resultObject( $this->db->query( $this->getQuery( $this->filter( $term ), true ) ) );
  95                 return new MySQLSearchResultSet( $resultSet, $this->searchTerms );
  96         }
  97
  98         /**
  99          * Perform a title-only search query and return a result set.
 100          *
 101          * @param string $term - Raw search term
 102          * @return MySQLSearchResultSet
 103          * @access public
 104          */
 105         function searchTitle( $term ) {
 106                 $resultSet = $this->db->resultObject( $this->db->query( $this->getQuery( $this->filter( $term ), false ) ) );
 107                 return new MySQLSearchResultSet( $resultSet, $this->searchTerms );
 108         }
 109
 110
 111         /**
 112          * Return a partial WHERE clause to exclude redirects, if so set
 113          * @return string
 114          * @private
 115          */
 116         function queryRedirect() {
 117                 if( $this->showRedirects ) {
 118                         return '';
 119                 } else {
 120                         return 'AND page_is_redirect=0';
 121                 }
 122         }
 123
 124         /**
 125          * Return a partial WHERE clause to limit the search to the given namespaces
 126          * @return string
 127          * @private
 128          */
 129         function queryNamespaces() {
 130                 if( is_null($this->namespaces) )
 131                         return '';  # search all
 132                 if ( !count( $this->namespaces ) ) {
 133                         $namespaces = '0';
 134                 } else {
 135                         $namespaces = $this->db->makeList( $this->namespaces );
 136                 }
 137                 return 'AND page_namespace IN (' . $namespaces . ')';
 138         }
 139
 140         /**
 141          * Return a LIMIT clause to limit results on the query.
 142          * @return string
 143          * @private
 144          */
 145         function queryLimit() {
 146                 return $this->db->limitResult( '', $this->limit, $this->offset );
 147         }
 148
 149         /**
 150          * Does not do anything for generic search engine
 151          * subclasses may define this though
 152          * @return string
 153          * @private
 154          */
 155         function queryRanking( $filteredTerm, $fulltext ) {
 156                 return '';
 157         }
 158
 159         /**
 160          * Construct the full SQL query to do the search.
 161          * The guts shoulds be constructed in queryMain()
 162          * @param string $filteredTerm
 163          * @param bool $fulltext
 164          * @private
 165          */
 166         function getQuery( $filteredTerm, $fulltext ) {
 167                 return $this->queryMain( $filteredTerm, $fulltext ) . ' ' .
 168                         $this->queryRedirect() . ' ' .
 169                         $this->queryNamespaces() . ' ' .
 170                         $this->queryRanking( $filteredTerm, $fulltext ) . ' ' .
 171                         $this->queryLimit();
 172         }
 173
 174
 175         /**
 176          * Picks which field to index on, depending on what type of query.
 177          * @param bool $fulltext
 178          * @return string
 179          */
 180         function getIndexField( $fulltext ) {
 181                 return $fulltext ? 'si_text' : 'si_title';
 182         }
 183
 184         /**
 185          * Get the base part of the search query.
 186          * The actual match syntax will depend on the server
 187          * version; MySQL 3 and MySQL 4 have different capabilities
 188          * in their fulltext search indexes.
 189          *
 190          * @param string $filteredTerm
 191          * @param bool $fulltext
 192          * @return string
 193          * @private
 194          */
 195         function queryMain( $filteredTerm, $fulltext ) {
 196                 $match = $this->parseQuery( $filteredTerm, $fulltext );
 197                 $page        = $this->db->tableName( 'page' );
 198                 $searchindex = $this->db->tableName( 'searchindex' );
 199                 return 'SELECT page_id, page_namespace, page_title ' .
 200                         "FROM $page,$searchindex " .
 201                         'WHERE page_id=si_page AND ' . $match;
 202         }
 203
 204         /**
 205          * Create or update the search index record for the given page.
 206          * Title and text should be pre-processed.
 207          *
 208          * @param int $id
 209          * @param string $title
 210          * @param string $text
 211          */
 212         function update( $id, $title, $text ) {
 213                 $dbw = wfGetDB( DB_MASTER );
 214                 $dbw->replace( 'searchindex',
 215                         array( 'si_page' ),
 216                         array(
 217                                 'si_page' => $id,
 218                                 'si_title' => $title,
 219                                 'si_text' => $text
 220                         ), __METHOD__ );
 221         }
 222
 223         /**
 224          * Update a search index record's title only.
 225          * Title should be pre-processed.
 226          *
 227          * @param int $id
 228          * @param string $title
 229          */
 230     function updateTitle( $id, $title ) {
 231                 $dbw = wfGetDB( DB_MASTER );
 232
 233                 $dbw->update( 'searchindex',
 234                         array( 'si_title' => $title ),
 235                         array( 'si_page'  => $id ),
 236                         __METHOD__,
 237                         array( $dbw->lowPriorityOption() ) );
 238         }
 239 }
 240
 241 /**
 242  * @ingroup Search
 243  */
 244 class MySQLSearchResultSet extends SearchResultSet {
 245         function MySQLSearchResultSet( $resultSet, $terms ) {
 246                 $this->mResultSet = $resultSet;
 247                 $this->mTerms = $terms;
 248         }
 249
 250         function termMatches() {
 251                 return $this->mTerms;
 252         }
 253
 254         function numRows() {
 255                 return $this->mResultSet->numRows();
 256         }
 257
 258         function next() {
 259                 $row = $this->mResultSet->fetchObject();
 260                 if( $row === false ) {
 261                         return false;
 262                 } else {
 263                         return new SearchResult( $row );
 264                 }
 265         }
 266
 267         function free() {
 268                 $this->mResultSet->free();
 269         }
 270 }