includes/search/SearchMssql.php

   1 <?php
   2 # Copyright (C) 2004 Brion Vibber <brion@pobox.com>
   3 # http://www.mediawiki.org/
   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 /**
  21  * Search engine hook base class for Mssql (ConText).
  22  * @addtogroup Search
  23  */
  24 class SearchMssql extends SearchEngine {
  25
  26         function __construct( $db ) {
  27                 $this->db = $db;
  28         }
  29
  30         /**
  31          * Perform a full text search query and return a result set.
  32          *
  33          * @param string $term - Raw search term
  34          * @return MssqlSearchResultSet
  35          * @access public
  36          */
  37         function searchText( $term ) {
  38                 $resultSet = $this->db->resultObject( $this->db->query( $this->getQuery( $this->filter( $term ), true ) ) );
  39                 return new MssqlSearchResultSet( $resultSet, $this->searchTerms );
  40         }
  41
  42         /**
  43          * Perform a title-only search query and return a result set.
  44          *
  45          * @param string $term - Raw search term
  46          * @return MssqlSearchResultSet
  47          * @access public
  48          */
  49         function searchTitle( $term ) {
  50                 $resultSet = $this->db->resultObject( $this->db->query( $this->getQuery( $this->filter( $term ), false ) ) );
  51                 return new MssqlSearchResultSet( $resultSet, $this->searchTerms );
  52         }
  53
  54
  55         /**
  56          * Return a partial WHERE clause to exclude redirects, if so set
  57          * @return string
  58          * @private
  59          */
  60         function queryRedirect() {
  61                 if ( $this->showRedirects ) {
  62                         return '';
  63                 } else {
  64                         return 'AND page_is_redirect=0';
  65                 }
  66         }
  67
  68         /**
  69          * Return a partial WHERE clause to limit the search to the given namespaces
  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          * @return string
  84          * @private
  85          */
  86         function queryLimit( $sql ) {
  87                 return $this->db->limitResult( $sql, $this->limit, $this->offset );
  88         }
  89
  90         /**
  91          * Does not do anything for generic search engine
  92          * subclasses may define this though
  93          * @return string
  94          * @private
  95          */
  96         function queryRanking( $filteredTerm, $fulltext ) {
  97                 return ' ORDER BY ftindex.[RANK] DESC'; // return ' ORDER BY score(1)';
  98         }
  99
 100         /**
 101          * Construct the full SQL query to do the search.
 102          * The guts shoulds be constructed in queryMain()
 103          * @param string $filteredTerm
 104          * @param bool $fulltext
 105          * @private
 106          */
 107         function getQuery( $filteredTerm, $fulltext ) {
 108                 return $this->queryLimit( $this->queryMain( $filteredTerm, $fulltext ) . ' ' .
 109                         $this->queryRedirect() . ' ' .
 110                         $this->queryNamespaces() . ' ' .
 111                         $this->queryRanking( $filteredTerm, $fulltext ) . ' ' );
 112         }
 113
 114
 115         /**
 116          * Picks which field to index on, depending on what type of query.
 117          * @param bool $fulltext
 118          * @return string
 119          */
 120         function getIndexField( $fulltext ) {
 121                 return $fulltext ? 'si_text' : 'si_title';
 122         }
 123
 124         /**
 125          * Get the base part of the search query.
 126          *
 127          * @param string $filteredTerm
 128          * @param bool $fulltext
 129          * @return string
 130          * @private
 131          */
 132         function queryMain( $filteredTerm, $fulltext ) {
 133                 $match = $this->parseQuery( $filteredTerm, $fulltext );
 134                 $page        = $this->db->tableName( 'page' );
 135                 $searchindex = $this->db->tableName( 'searchindex' );
 136
 137                 return 'SELECT page_id, page_namespace, page_title, ftindex.[RANK]' .
 138                         "FROM $page,FREETEXTTABLE($searchindex , $match, LANGUAGE 'English') as ftindex " .
 139                         'WHERE page_id=ftindex.[KEY] ';
 140         }
 141
 142         /** @todo document */
 143         function parseQuery( $filteredText, $fulltext ) {
 144                 global $wgContLang;
 145                 $lc = SearchEngine::legalSearchChars();
 146                 $this->searchTerms = array();
 147
 148                 # FIXME: This doesn't handle parenthetical expressions.
 149                 $m = array();
 150                 $q = array();
 151
 152                 if ( preg_match_all( '/([-+<>~]?)(([' . $lc . ']+)(\*?)|"[^"]*")/',
 153                         $filteredText, $m, PREG_SET_ORDER ) ) {
 154                         foreach ( $m as $terms ) {
 155                                 $q[] = $terms[1] . $wgContLang->normalizeForSearch( $terms[2] );
 156
 157                                 if ( !empty( $terms[3] ) ) {
 158                                         $regexp = preg_quote( $terms[3], '/' );
 159                                         if ( $terms[4] )
 160                                                 $regexp .= "[0-9A-Za-z_]+";
 161                                 } else {
 162                                         $regexp = preg_quote( str_replace( '"', '', $terms[2] ), '/' );
 163                                 }
 164                                 $this->searchTerms[] = $regexp;
 165                         }
 166                 }
 167
 168                 $searchon = $this->db->strencode( join( ',', $q ) );
 169                 $field = $this->getIndexField( $fulltext );
 170                 return "$field, '$searchon'";
 171         }
 172
 173         /**
 174          * Create or update the search index record for the given page.
 175          * Title and text should be pre-processed.
 176          *
 177          * @param int $id
 178          * @param string $title
 179          * @param string $text
 180          */
 181         function update( $id, $title, $text ) {
 182                 // We store the column data as UTF-8 byte order marked binary stream
 183                 // because we are invoking the plain text IFilter on it so that, and we want it
 184                 // to properly decode the stream as UTF-8.  SQL doesn't support UTF8 as a data type
 185                 // but the indexer will correctly handle it by this method.  Since all we are doing
 186                 // is passing this data to the indexer and never retrieving it via PHP, this will save space
 187                 $table = $this->db->tableName( 'searchindex' );
 188                 $utf8bom = '0xEFBBBF';
 189                 $si_title = $utf8bom . bin2hex( $title );
 190                 $si_text = $utf8bom . bin2hex( $text );
 191                 $sql = "DELETE FROM $table WHERE si_page = $id;";
 192                 $sql .= "INSERT INTO $table (si_page, si_title, si_text) VALUES ($id, $si_title, $si_text)";
 193                 return $this->db->query( $sql, 'SearchMssql::update' );
 194         }
 195
 196         /**
 197          * Update a search index record's title only.
 198          * Title should be pre-processed.
 199          *
 200          * @param int $id
 201          * @param string $title
 202          */
 203         function updateTitle( $id, $title ) {
 204                 $table = $this->db->tableName( 'searchindex' );
 205
 206                 // see update for why we are using the utf8bom
 207                 $utf8bom = '0xEFBBBF';
 208                 $si_title = $utf8bom . bin2hex( $title );
 209                 $sql = "DELETE FROM $table WHERE si_page = $id;";
 210                 $sql .= "INSERT INTO $table (si_page, si_title, si_text) VALUES ($id, $si_title, 0x00)";
 211                 return $this->db->query( $sql, 'SearchMssql::updateTitle' );
 212         }
 213 }
 214
 215 /**
 216  * @addtogroup Search
 217  */
 218 class MssqlSearchResultSet extends SearchResultSet {
 219         function __construct( $resultSet, $terms ) {
 220                 $this->mResultSet = $resultSet;
 221                 $this->mTerms = $terms;
 222         }
 223
 224         function termMatches() {
 225                 return $this->mTerms;
 226         }
 227
 228         function numRows() {
 229                 return $this->mResultSet->numRows();
 230         }
 231
 232         function next() {
 233                 $row = $this->mResultSet->fetchObject();
 234                 if ( $row === false )
 235                         return false;
 236                 return new SearchResult( $row );
 237         }
 238 }
 239
 240