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 Wikimedia\Rdbms\IResultWrapper;
  25
  26 /**
  27  * Search engine hook base class for Mssql (ConText).
  28  * @ingroup Search
  29  */
  30 class SearchMssql extends SearchDatabase {
  31         /**
  32          * Perform a full text search query and return a result set.
  33          *
  34          * @param string $term Raw search term
  35          * @return SqlSearchResultSet
  36          */
  37         protected function doSearchTextInDB( $term ) {
  38                 $resultSet = $this->db->query( $this->getQuery( $this->filter( $term ), true ) );
  39                 return new SqlSearchResultSet( $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 SqlSearchResultSet
  47          */
  48         protected function doSearchTitleInDB( $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          */
  58         private function queryNamespaces() {
  59                 $namespaces = implode( ',', $this->namespaces );
  60                 if ( $namespaces == '' ) {
  61                         $namespaces = '0';
  62                 }
  63                 return 'AND page_namespace IN (' . $namespaces . ')';
  64         }
  65
  66         /**
  67          * Return a LIMIT clause to limit results on the query.
  68          *
  69          * @param string $sql
  70          *
  71          * @return string
  72          */
  73         private function queryLimit( $sql ) {
  74                 return $this->db->limitResult( $sql, $this->limit, $this->offset );
  75         }
  76
  77         /**
  78          * Does not do anything for generic search engine
  79          * subclasses may define this though
  80          *
  81          * @param string $filteredTerm
  82          * @param bool $fulltext
  83          * @return string
  84          */
  85         function queryRanking( $filteredTerm, $fulltext ) {
  86                 return ' ORDER BY ftindex.[RANK] DESC'; // return ' ORDER BY score(1)';
  87         }
  88
  89         /**
  90          * Construct the full SQL query to do the search.
  91          * The guts shoulds be constructed in queryMain()
  92          *
  93          * @param string $filteredTerm
  94          * @param bool $fulltext
  95          * @return string
  96          */
  97         private function getQuery( $filteredTerm, $fulltext ) {
  98                 return $this->queryLimit( $this->queryMain( $filteredTerm, $fulltext ) . ' ' .
  99                         $this->queryNamespaces() . ' ' .
 100                         $this->queryRanking( $filteredTerm, $fulltext ) . ' ' );
 101         }
 102
 103         /**
 104          * Picks which field to index on, depending on what type of query.
 105          *
 106          * @param bool $fulltext
 107          * @return string
 108          */
 109         function getIndexField( $fulltext ) {
 110                 return $fulltext ? 'si_text' : 'si_title';
 111         }
 112
 113         /**
 114          * Get the base part of the search query.
 115          *
 116          * @param string $filteredTerm
 117          * @param bool $fulltext
 118          * @return string
 119          */
 120         private function queryMain( $filteredTerm, $fulltext ) {
 121                 $match = $this->parseQuery( $filteredTerm, $fulltext );
 122                 $page = $this->db->tableName( 'page' );
 123                 $searchindex = $this->db->tableName( 'searchindex' );
 124
 125                 return 'SELECT page_id, page_namespace, page_title, ftindex.[RANK]' .
 126                         "FROM $page,FREETEXTTABLE($searchindex , $match, LANGUAGE 'English') as ftindex " .
 127                         'WHERE page_id=ftindex.[KEY] ';
 128         }
 129
 130         /** @todo document
 131          * @param string $filteredText
 132          * @param bool $fulltext
 133          * @return string
 134          */
 135         private function parseQuery( $filteredText, $fulltext ) {
 136                 global $wgContLang;
 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] . $wgContLang->normalizeForSearch( $terms[2] );
 148
 149                                 if ( !empty( $terms[3] ) ) {
 150                                         $regexp = preg_quote( $terms[3], '/' );
 151                                         if ( $terms[4] ) {
 152                                                 $regexp .= "[0-9A-Za-z_]+";
 153                                         }
 154                                 } else {
 155                                         $regexp = preg_quote( str_replace( '"', '', $terms[2] ), '/' );
 156                                 }
 157                                 $this->searchTerms[] = $regexp;
 158                         }
 159                 }
 160
 161                 $searchon = $this->db->addQuotes( implode( ',', $q ) );
 162                 $field = $this->getIndexField( $fulltext );
 163                 return "$field, $searchon";
 164         }
 165
 166         /**
 167          * Create or update the search index record for the given page.
 168          * Title and text should be pre-processed.
 169          *
 170          * @param int $id
 171          * @param string $title
 172          * @param string $text
 173          * @return bool|IResultWrapper
 174          */
 175         function update( $id, $title, $text ) {
 176                 // We store the column data as UTF-8 byte order marked binary stream
 177                 // because we are invoking the plain text IFilter on it so that, and we want it
 178                 // to properly decode the stream as UTF-8.  SQL doesn't support UTF8 as a data type
 179                 // but the indexer will correctly handle it by this method.  Since all we are doing
 180                 // is passing this data to the indexer and never retrieving it via PHP, this will save space
 181                 $table = $this->db->tableName( 'searchindex' );
 182                 $utf8bom = '0xEFBBBF';
 183                 $si_title = $utf8bom . bin2hex( $title );
 184                 $si_text = $utf8bom . bin2hex( $text );
 185                 $sql = "DELETE FROM $table WHERE si_page = $id;";
 186                 $sql .= "INSERT INTO $table (si_page, si_title, si_text) VALUES ($id, $si_title, $si_text)";
 187                 return $this->db->query( $sql, 'SearchMssql::update' );
 188         }
 189
 190         /**
 191          * Update a search index record's title only.
 192          * Title should be pre-processed.
 193          *
 194          * @param int $id
 195          * @param string $title
 196          * @return bool|IResultWrapper
 197          */
 198         function updateTitle( $id, $title ) {
 199                 $table = $this->db->tableName( 'searchindex' );
 200
 201                 // see update for why we are using the utf8bom
 202                 $utf8bom = '0xEFBBBF';
 203                 $si_title = $utf8bom . bin2hex( $title );
 204                 $sql = "DELETE FROM $table WHERE si_page = $id;";
 205                 $sql .= "INSERT INTO $table (si_page, si_title, si_text) VALUES ($id, $si_title, 0x00)";
 206                 return $this->db->query( $sql, 'SearchMssql::updateTitle' );
 207         }
 208 }