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] ) $regexp .= "[0-9A-Za-z_]+";
  58                                 } else {
  59                                         // Match the quoted term in result highlighting...
  60                                         $regexp = preg_quote( str_replace( '"', '', $terms[2] ), '/' );
  61                                 }
  62                                 $this->searchTerms[] = $regexp;
  63                         }
  64                         wfDebug( "Would search with '$searchon'\n" );
  65                         wfDebug( 'Match with /' . implode( '|', $this->searchTerms ) . "/\n" );
  66                 } else {
  67                         wfDebug( "Can't understand search query '{$filteredText}'\n" );
  68                 }
  69
  70                 $searchon = $this->db->strencode( $searchon );
  71                 $field = $this->getIndexField( $fulltext );
  72                 return " MATCH($field) AGAINST('$searchon' IN BOOLEAN MODE) ";
  73         }
  74
  75         public static function legalSearchChars() {
  76                 return "\"*" . parent::legalSearchChars();
  77         }
  78
  79         /**
  80          * Perform a full text search query and return a result set.
  81          *
  82          * @param string $term - Raw search term
  83          * @return MySQLSearchResultSet
  84          * @access public
  85          */
  86         function searchText( $term ) {
  87                 $resultSet = $this->db->resultObject( $this->db->query( $this->getQuery( $this->filter( $term ), true ) ) );
  88                 return new MySQLSearchResultSet( $resultSet, $this->searchTerms );
  89         }
  90
  91         /**
  92          * Perform a title-only search query and return a result set.
  93          *
  94          * @param string $term - Raw search term
  95          * @return MySQLSearchResultSet
  96          * @access public
  97          */
  98         function searchTitle( $term ) {
  99                 $resultSet = $this->db->resultObject( $this->db->query( $this->getQuery( $this->filter( $term ), false ) ) );
 100                 return new MySQLSearchResultSet( $resultSet, $this->searchTerms );
 101         }
 102
 103
 104         /**
 105          * Return a partial WHERE clause to exclude redirects, if so set
 106          * @return string
 107          * @private
 108          */
 109         function queryRedirect() {
 110                 if( $this->showRedirects ) {
 111                         return '';
 112                 } else {
 113                         return 'AND page_is_redirect=0';
 114                 }
 115         }
 116
 117         /**
 118          * Return a partial WHERE clause to limit the search to the given namespaces
 119          * @return string
 120          * @private
 121          */
 122         function queryNamespaces() {
 123                 if( is_null($this->namespaces) )
 124                         return '';  # search all
 125                 $namespaces = implode( ',', $this->namespaces );
 126                 if ($namespaces == '') {
 127                         $namespaces = '0';
 128                 }
 129                 return 'AND page_namespace IN (' . $namespaces . ')';
 130         }
 131
 132         /**
 133          * Return a LIMIT clause to limit results on the query.
 134          * @return string
 135          * @private
 136          */
 137         function queryLimit() {
 138                 return $this->db->limitResult( '', $this->limit, $this->offset );
 139         }
 140
 141         /**
 142          * Does not do anything for generic search engine
 143          * subclasses may define this though
 144          * @return string
 145          * @private
 146          */
 147         function queryRanking( $filteredTerm, $fulltext ) {
 148                 return '';
 149         }
 150
 151         /**
 152          * Construct the full SQL query to do the search.
 153          * The guts shoulds be constructed in queryMain()
 154          * @param string $filteredTerm
 155          * @param bool $fulltext
 156          * @private
 157          */
 158         function getQuery( $filteredTerm, $fulltext ) {
 159                 return $this->queryMain( $filteredTerm, $fulltext ) . ' ' .
 160                         $this->queryRedirect() . ' ' .
 161                         $this->queryNamespaces() . ' ' .
 162                         $this->queryRanking( $filteredTerm, $fulltext ) . ' ' .
 163                         $this->queryLimit();
 164         }
 165
 166
 167         /**
 168          * Picks which field to index on, depending on what type of query.
 169          * @param bool $fulltext
 170          * @return string
 171          */
 172         function getIndexField( $fulltext ) {
 173                 return $fulltext ? 'si_text' : 'si_title';
 174         }
 175
 176         /**
 177          * Get the base part of the search query.
 178          * The actual match syntax will depend on the server
 179          * version; MySQL 3 and MySQL 4 have different capabilities
 180          * in their fulltext search indexes.
 181          *
 182          * @param string $filteredTerm
 183          * @param bool $fulltext
 184          * @return string
 185          * @private
 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 int $id
 201          * @param string $title
 202          * @param string $text
 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 int $id
 220          * @param string $title
 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 }