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