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                 $dbr = $this->lb->getConnectionRef( DB_REPLICA );
  40                 $resultSet = $dbr->query( $this->getQuery( $this->filter( $term ), true ) );
  41
  42                 return new SqlSearchResultSet( $resultSet, $this->searchTerms );
  43         }
  44
  45         /**
  46          * Perform a title-only search query and return a result set.
  47          *
  48          * @param string $term Raw search term
  49          * @return SqlSearchResultSet
  50          */
  51         protected function doSearchTitleInDB( $term ) {
  52                 $dbr = $this->lb->getConnectionRef( DB_REPLICA );
  53                 $resultSet = $dbr->query( $this->getQuery( $this->filter( $term ), false ) );
  54
  55                 return new SqlSearchResultSet( $resultSet, $this->searchTerms );
  56         }
  57
  58         /**
  59          * Return a partial WHERE clause to limit the search to the given namespaces
  60          *
  61          * @return string
  62          */
  63         private function queryNamespaces() {
  64                 $namespaces = implode( ',', $this->namespaces );
  65                 if ( $namespaces == '' ) {
  66                         $namespaces = '0';
  67                 }
  68                 return 'AND page_namespace IN (' . $namespaces . ')';
  69         }
  70
  71         /**
  72          * Return a LIMIT clause to limit results on the query.
  73          *
  74          * @param string $sql
  75          *
  76          * @return string
  77          */
  78         private function queryLimit( $sql ) {
  79                 $dbr = $this->lb->getConnectionRef( DB_REPLICA );
  80
  81                 return $dbr->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          * @param string $filteredTerm
  89          * @param bool $fulltext
  90          * @return string
  91          */
  92         function queryRanking( $filteredTerm, $fulltext ) {
  93                 return ' ORDER BY ftindex.[RANK] DESC'; // return ' ORDER BY score(1)';
  94         }
  95
  96         /**
  97          * Construct the full SQL query to do the search.
  98          * The guts shoulds be constructed in queryMain()
  99          *
 100          * @param string $filteredTerm
 101          * @param bool $fulltext
 102          * @return string
 103          */
 104         private function getQuery( $filteredTerm, $fulltext ) {
 105                 return $this->queryLimit( $this->queryMain( $filteredTerm, $fulltext ) . ' ' .
 106                         $this->queryNamespaces() . ' ' .
 107                         $this->queryRanking( $filteredTerm, $fulltext ) . ' ' );
 108         }
 109
 110         /**
 111          * Picks which field to index on, depending on what type of query.
 112          *
 113          * @param bool $fulltext
 114          * @return string
 115          */
 116         function getIndexField( $fulltext ) {
 117                 return $fulltext ? 'si_text' : 'si_title';
 118         }
 119
 120         /**
 121          * Get the base part of the search query.
 122          *
 123          * @param string $filteredTerm
 124          * @param bool $fulltext
 125          * @return string
 126          */
 127         private function queryMain( $filteredTerm, $fulltext ) {
 128                 $match = $this->parseQuery( $filteredTerm, $fulltext );
 129                 $dbr = $this->lb->getMaintenanceConnectionRef( DB_REPLICA );
 130                 $page = $dbr->tableName( 'page' );
 131                 $searchindex = $dbr->tableName( 'searchindex' );
 132
 133                 return 'SELECT page_id, page_namespace, page_title, ftindex.[RANK]' .
 134                         "FROM $page,FREETEXTTABLE($searchindex , $match, LANGUAGE 'English') as ftindex " .
 135                         'WHERE page_id=ftindex.[KEY] ';
 136         }
 137
 138         /** @todo document
 139          * @param string $filteredText
 140          * @param bool $fulltext
 141          * @return string
 142          */
 143         private function parseQuery( $filteredText, $fulltext ) {
 144                 $lc = $this->legalSearchChars( self::CHARS_NO_SYNTAX );
 145                 $this->searchTerms = [];
 146
 147                 # @todo FIXME: This doesn't handle parenthetical expressions.
 148                 $m = [];
 149                 $q = [];
 150
 151                 if ( preg_match_all( '/([-+<>~]?)(([' . $lc . ']+)(\*?)|"[^"]*")/',
 152                         $filteredText, $m, PREG_SET_ORDER ) ) {
 153                         foreach ( $m as $terms ) {
 154                                 $q[] = $terms[1] . MediaWikiServices::getInstance()->getContentLanguage()->
 155                                         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                                         }
 162                                 } else {
 163                                         $regexp = preg_quote( str_replace( '"', '', $terms[2] ), '/' );
 164                                 }
 165                                 $this->searchTerms[] = $regexp;
 166                         }
 167                 }
 168
 169                 $dbr = $this->lb->getConnectionRef( DB_REPLICA );
 170                 $searchon = $dbr->addQuotes( implode( ',', $q ) );
 171                 $field = $this->getIndexField( $fulltext );
 172
 173                 return "$field, $searchon";
 174         }
 175
 176         /**
 177          * Create or update the search index record for the given page.
 178          * Title and text should be pre-processed.
 179          *
 180          * @param int $id
 181          * @param string $title
 182          * @param string $text
 183          * @return bool|IResultWrapper
 184          */
 185         function update( $id, $title, $text ) {
 186                 // We store the column data as UTF-8 byte order marked binary stream
 187                 // because we are invoking the plain text IFilter on it so that, and we want it
 188                 // to properly decode the stream as UTF-8.  SQL doesn't support UTF8 as a data type
 189                 // but the indexer will correctly handle it by this method.  Since all we are doing
 190                 // is passing this data to the indexer and never retrieving it via PHP, this will save space
 191                 $dbr = $this->lb->getMaintenanceConnectionRef( DB_MASTER );
 192                 $table = $dbr->tableName( 'searchindex' );
 193                 $utf8bom = '0xEFBBBF';
 194                 $si_title = $utf8bom . bin2hex( $title );
 195                 $si_text = $utf8bom . bin2hex( $text );
 196                 $sql = "DELETE FROM $table WHERE si_page = $id;";
 197                 $sql .= "INSERT INTO $table (si_page, si_title, si_text) VALUES ($id, $si_title, $si_text)";
 198                 return $dbr->query( $sql, 'SearchMssql::update' );
 199         }
 200
 201         /**
 202          * Update a search index record's title only.
 203          * Title should be pre-processed.
 204          *
 205          * @param int $id
 206          * @param string $title
 207          * @return bool|IResultWrapper
 208          */
 209         function updateTitle( $id, $title ) {
 210                 $dbr = $this->lb->getMaintenanceConnectionRef( DB_MASTER );
 211                 $table = $dbr->tableName( 'searchindex' );
 212
 213                 // see update for why we are using the utf8bom
 214                 $utf8bom = '0xEFBBBF';
 215                 $si_title = $utf8bom . bin2hex( $title );
 216                 $sql = "DELETE FROM $table WHERE si_page = $id;";
 217                 $sql .= "INSERT INTO $table (si_page, si_title, si_text) VALUES ($id, $si_title, 0x00)";
 218                 return $dbr->query( $sql, 'SearchMssql::updateTitle' );
 219         }
 220 }