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