includes/search/SearchMssql.php

   1 <?php
   2 /**
   3  * Mssql search engine
   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  * @file
  21  * @ingroup Search
  22  */
  23
  24 /**
  25  * Search engine hook base class for Mssql (ConText).
  26  * @ingroup Search
  27  */
  28 class SearchMssql extends SearchEngine {
  29
  30         /**
  31          * Creates an instance of this class
  32          * @param $db DatabaseMssql: database object
  33          */
  34         function __construct( $db ) {
  35                 parent::__construct( $db );
  36         }
  37
  38         /**
  39          * Perform a full text search query and return a result set.
  40          *
  41          * @param $term String: raw search term
  42          * @return MssqlSearchResultSet
  43          * @access public
  44          */
  45         function searchText( $term ) {
  46                 $resultSet = $this->db->resultObject( $this->db->query( $this->getQuery( $this->filter( $term ), true ) ) );
  47                 return new MssqlSearchResultSet( $resultSet, $this->searchTerms );
  48         }
  49
  50         /**
  51          * Perform a title-only search query and return a result set.
  52          *
  53          * @param $term String: raw search term
  54          * @return MssqlSearchResultSet
  55          * @access public
  56          */
  57         function searchTitle( $term ) {
  58                 $resultSet = $this->db->resultObject( $this->db->query( $this->getQuery( $this->filter( $term ), false ) ) );
  59                 return new MssqlSearchResultSet( $resultSet, $this->searchTerms );
  60         }
  61
  62
  63         /**
  64          * Return a partial WHERE clause to exclude redirects, if so set
  65          *
  66          * @return String
  67          * @private
  68          */
  69         function queryRedirect() {
  70                 if ( $this->showRedirects ) {
  71                         return '';
  72                 } else {
  73                         return 'AND page_is_redirect=0';
  74                 }
  75         }
  76
  77         /**
  78          * Return a partial WHERE clause to limit the search to the given namespaces
  79          *
  80          * @return String
  81          * @private
  82          */
  83         function queryNamespaces() {
  84                 $namespaces = implode( ',', $this->namespaces );
  85                 if ( $namespaces == '' ) {
  86                         $namespaces = '0';
  87                 }
  88                 return 'AND page_namespace IN (' . $namespaces . ')';
  89         }
  90
  91         /**
  92          * Return a LIMIT clause to limit results on the query.
  93          *
  94          * @param $sql string
  95          *
  96          * @return String
  97          */
  98         function queryLimit( $sql ) {
  99                 return $this->db->limitResult( $sql, $this->limit, $this->offset );
 100         }
 101
 102         /**
 103          * Does not do anything for generic search engine
 104          * subclasses may define this though
 105          *
 106          * @return String
 107          */
 108         function queryRanking( $filteredTerm, $fulltext ) {
 109                 return ' ORDER BY ftindex.[RANK] DESC'; // return ' ORDER BY score(1)';
 110         }
 111
 112         /**
 113          * Construct the full SQL query to do the search.
 114          * The guts shoulds be constructed in queryMain()
 115          *
 116          * @param $filteredTerm String
 117          * @param $fulltext Boolean
 118          * @return String
 119          */
 120         function getQuery( $filteredTerm, $fulltext ) {
 121                 return $this->queryLimit( $this->queryMain( $filteredTerm, $fulltext ) . ' ' .
 122                         $this->queryRedirect() . ' ' .
 123                         $this->queryNamespaces() . ' ' .
 124                         $this->queryRanking( $filteredTerm, $fulltext ) . ' ' );
 125         }
 126
 127         /**
 128          * Picks which field to index on, depending on what type of query.
 129          *
 130          * @param $fulltext Boolean
 131          * @return string
 132          */
 133         function getIndexField( $fulltext ) {
 134                 return $fulltext ? 'si_text' : 'si_title';
 135         }
 136
 137         /**
 138          * Get the base part of the search query.
 139          *
 140          * @param $filteredTerm String
 141          * @param $fulltext Boolean
 142          * @return String
 143          * @private
 144          */
 145         function queryMain( $filteredTerm, $fulltext ) {
 146                 $match = $this->parseQuery( $filteredTerm, $fulltext );
 147                 $page        = $this->db->tableName( 'page' );
 148                 $searchindex = $this->db->tableName( 'searchindex' );
 149
 150                 return 'SELECT page_id, page_namespace, page_title, ftindex.[RANK]' .
 151                         "FROM $page,FREETEXTTABLE($searchindex , $match, LANGUAGE 'English') as ftindex " .
 152                         'WHERE page_id=ftindex.[KEY] ';
 153         }
 154
 155         /** @todo document
 156          * @return string
 157          */
 158         function parseQuery( $filteredText, $fulltext ) {
 159                 global $wgContLang;
 160                 $lc = SearchEngine::legalSearchChars();
 161                 $this->searchTerms = array();
 162
 163                 # @todo FIXME: This doesn't handle parenthetical expressions.
 164                 $m = array();
 165                 $q = array();
 166
 167                 if ( preg_match_all( '/([-+<>~]?)(([' . $lc . ']+)(\*?)|"[^"]*")/',
 168                         $filteredText, $m, PREG_SET_ORDER ) ) {
 169                         foreach ( $m as $terms ) {
 170                                 $q[] = $terms[1] . $wgContLang->normalizeForSearch( $terms[2] );
 171
 172                                 if ( !empty( $terms[3] ) ) {
 173                                         $regexp = preg_quote( $terms[3], '/' );
 174                                         if ( $terms[4] )
 175                                                 $regexp .= "[0-9A-Za-z_]+";
 176                                 } else {
 177                                         $regexp = preg_quote( str_replace( '"', '', $terms[2] ), '/' );
 178                                 }
 179                                 $this->searchTerms[] = $regexp;
 180                         }
 181                 }
 182
 183                 $searchon = $this->db->strencode( join( ',', $q ) );
 184                 $field = $this->getIndexField( $fulltext );
 185                 return "$field, '$searchon'";
 186         }
 187
 188         /**
 189          * Create or update the search index record for the given page.
 190          * Title and text should be pre-processed.
 191          *
 192          * @param $id Integer
 193          * @param $title String
 194          * @param $text String
 195          * @return bool|ResultWrapper
 196          */
 197         function update( $id, $title, $text ) {
 198                 // We store the column data as UTF-8 byte order marked binary stream
 199                 // because we are invoking the plain text IFilter on it so that, and we want it
 200                 // to properly decode the stream as UTF-8.  SQL doesn't support UTF8 as a data type
 201                 // but the indexer will correctly handle it by this method.  Since all we are doing
 202                 // is passing this data to the indexer and never retrieving it via PHP, this will save space
 203                 $table = $this->db->tableName( 'searchindex' );
 204                 $utf8bom = '0xEFBBBF';
 205                 $si_title = $utf8bom . bin2hex( $title );
 206                 $si_text = $utf8bom . bin2hex( $text );
 207                 $sql = "DELETE FROM $table WHERE si_page = $id;";
 208                 $sql .= "INSERT INTO $table (si_page, si_title, si_text) VALUES ($id, $si_title, $si_text)";
 209                 return $this->db->query( $sql, 'SearchMssql::update' );
 210         }
 211
 212         /**
 213          * Update a search index record's title only.
 214          * Title should be pre-processed.
 215          *
 216          * @param $id Integer
 217          * @param $title String
 218          * @return bool|ResultWrapper
 219          */
 220         function updateTitle( $id, $title ) {
 221                 $table = $this->db->tableName( 'searchindex' );
 222
 223                 // see update for why we are using the utf8bom
 224                 $utf8bom = '0xEFBBBF';
 225                 $si_title = $utf8bom . bin2hex( $title );
 226                 $sql = "DELETE FROM $table WHERE si_page = $id;";
 227                 $sql .= "INSERT INTO $table (si_page, si_title, si_text) VALUES ($id, $si_title, 0x00)";
 228                 return $this->db->query( $sql, 'SearchMssql::updateTitle' );
 229         }
 230 }
 231
 232 /**
 233  * @ingroup Search
 234  */
 235 class MssqlSearchResultSet extends SearchResultSet {
 236         function __construct( $resultSet, $terms ) {
 237                 $this->mResultSet = $resultSet;
 238                 $this->mTerms = $terms;
 239         }
 240
 241         function termMatches() {
 242                 return $this->mTerms;
 243         }
 244
 245         function numRows() {
 246                 return $this->mResultSet->numRows();
 247         }
 248
 249         function next() {
 250                 $row = $this->mResultSet->fetchObject();
 251                 if ( $row === false )
 252                         return false;
 253                 return new SearchResult( $row );
 254         }
 255 }