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 use MediaWiki\MediaWikiServices;
  25 use Wikimedia\Rdbms\IResultWrapper;
  26
  27 /**
  28  * Search engine hook base class for Mssql (ConText).
  29  * @ingroup Search
  30  */
  31 class SearchMssql extends SearchDatabase {
  32         /**
  33          * Perform a full text search query and return a result set.
  34          *
  35          * @param string $term Raw search term
  36          * @return SqlSearchResultSet
  37          */
  38         protected function doSearchTextInDB( $term ) {
  39                 $resultSet = $this->db->query( $this->getQuery( $this->filter( $term ), true ) );
  40                 return new SqlSearchResultSet( $resultSet, $this->searchTerms );
  41         }
  42
  43         /**
  44          * Perform a title-only search query and return a result set.
  45          *
  46          * @param string $term Raw search term
  47          * @return SqlSearchResultSet
  48          */
  49         protected function doSearchTitleInDB( $term ) {
  50                 $resultSet = $this->db->query( $this->getQuery( $this->filter( $term ), false ) );
  51                 return new SqlSearchResultSet( $resultSet, $this->searchTerms );
  52         }
  53
  54         /**
  55          * Return a partial WHERE clause to limit the search to the given namespaces
  56          *
  57          * @return string
  58          */
  59         private 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         private 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         private 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          */
 121         private function queryMain( $filteredTerm, $fulltext ) {
 122                 $match = $this->parseQuery( $filteredTerm, $fulltext );
 123                 $page = $this->db->tableName( 'page' );
 124                 $searchindex = $this->db->tableName( 'searchindex' );
 125
 126                 return 'SELECT page_id, page_namespace, page_title, ftindex.[RANK]' .
 127                         "FROM $page,FREETEXTTABLE($searchindex , $match, LANGUAGE 'English') as ftindex " .
 128                         'WHERE page_id=ftindex.[KEY] ';
 129         }
 130
 131         /** @todo document
 132          * @param string $filteredText
 133          * @param bool $fulltext
 134          * @return string
 135          */
 136         private function parseQuery( $filteredText, $fulltext ) {
 137                 $lc = $this->legalSearchChars( self::CHARS_NO_SYNTAX );
 138                 $this->searchTerms = [];
 139
 140                 # @todo FIXME: This doesn't handle parenthetical expressions.
 141                 $m = [];
 142                 $q = [];
 143
 144                 if ( preg_match_all( '/([-+<>~]?)(([' . $lc . ']+)(\*?)|"[^"]*")/',
 145                         $filteredText, $m, PREG_SET_ORDER ) ) {
 146                         foreach ( $m as $terms ) {
 147                                 $q[] = $terms[1] . MediaWikiServices::getInstance()->getContentLanguage()->
 148                                         normalizeForSearch( $terms[2] );
 149
 150                                 if ( !empty( $terms[3] ) ) {
 151                                         $regexp = preg_quote( $terms[3], '/' );
 152                                         if ( $terms[4] ) {
 153                                                 $regexp .= "[0-9A-Za-z_]+";
 154                                         }
 155                                 } else {
 156                                         $regexp = preg_quote( str_replace( '"', '', $terms[2] ), '/' );
 157                                 }
 158                                 $this->searchTerms[] = $regexp;
 159                         }
 160                 }
 161
 162                 $searchon = $this->db->addQuotes( implode( ',', $q ) );
 163                 $field = $this->getIndexField( $fulltext );
 164                 return "$field, $searchon";
 165         }
 166
 167         /**
 168          * Create or update the search index record for the given page.
 169          * Title and text should be pre-processed.
 170          *
 171          * @param int $id
 172          * @param string $title
 173          * @param string $text
 174          * @return bool|IResultWrapper
 175          */
 176         function update( $id, $title, $text ) {
 177                 // We store the column data as UTF-8 byte order marked binary stream
 178                 // because we are invoking the plain text IFilter on it so that, and we want it
 179                 // to properly decode the stream as UTF-8.  SQL doesn't support UTF8 as a data type
 180                 // but the indexer will correctly handle it by this method.  Since all we are doing
 181                 // is passing this data to the indexer and never retrieving it via PHP, this will save space
 182                 $table = $this->db->tableName( 'searchindex' );
 183                 $utf8bom = '0xEFBBBF';
 184                 $si_title = $utf8bom . bin2hex( $title );
 185                 $si_text = $utf8bom . bin2hex( $text );
 186                 $sql = "DELETE FROM $table WHERE si_page = $id;";
 187                 $sql .= "INSERT INTO $table (si_page, si_title, si_text) VALUES ($id, $si_title, $si_text)";
 188                 return $this->db->query( $sql, 'SearchMssql::update' );
 189         }
 190
 191         /**
 192          * Update a search index record's title only.
 193          * Title should be pre-processed.
 194          *
 195          * @param int $id
 196          * @param string $title
 197          * @return bool|IResultWrapper
 198          */
 199         function updateTitle( $id, $title ) {
 200                 $table = $this->db->tableName( 'searchindex' );
 201
 202                 // see update for why we are using the utf8bom
 203                 $utf8bom = '0xEFBBBF';
 204                 $si_title = $utf8bom . bin2hex( $title );
 205                 $sql = "DELETE FROM $table WHERE si_page = $id;";
 206                 $sql .= "INSERT INTO $table (si_page, si_title, si_text) VALUES ($id, $si_title, 0x00)";
 207                 return $this->db->query( $sql, 'SearchMssql::updateTitle' );
 208         }
 209 }