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