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