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