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          */
  35         protected function doSearchText( $term ) {
  36                 $resultSet = $this->db->query( $this->getQuery( $this->filter( $term ), true ) );
  37                 return new SqlSearchResultSet( $resultSet, $this->searchTerms );
  38         }
  39
  40         /**
  41          * Perform a title-only search query and return a result set.
  42          *
  43          * @param string $term Raw search term
  44          * @return SqlSearchResultSet
  45          */
  46         protected function doSearchTitle( $term ) {
  47                 $resultSet = $this->db->query( $this->getQuery( $this->filter( $term ), false ) );
  48                 return new SqlSearchResultSet( $resultSet, $this->searchTerms );
  49         }
  50
  51         /**
  52          * Return a partial WHERE clause to limit the search to the given namespaces
  53          *
  54          * @return string
  55          */
  56         private function queryNamespaces() {
  57                 $namespaces = implode( ',', $this->namespaces );
  58                 if ( $namespaces == '' ) {
  59                         $namespaces = '0';
  60                 }
  61                 return 'AND page_namespace IN (' . $namespaces . ')';
  62         }
  63
  64         /**
  65          * Return a LIMIT clause to limit results on the query.
  66          *
  67          * @param string $sql
  68          *
  69          * @return string
  70          */
  71         private function queryLimit( $sql ) {
  72                 return $this->db->limitResult( $sql, $this->limit, $this->offset );
  73         }
  74
  75         /**
  76          * Does not do anything for generic search engine
  77          * subclasses may define this though
  78          *
  79          * @param string $filteredTerm
  80          * @param bool $fulltext
  81          * @return string
  82          */
  83         function queryRanking( $filteredTerm, $fulltext ) {
  84                 return ' ORDER BY ftindex.[RANK] DESC'; // return ' ORDER BY score(1)';
  85         }
  86
  87         /**
  88          * Construct the full SQL query to do the search.
  89          * The guts shoulds be constructed in queryMain()
  90          *
  91          * @param string $filteredTerm
  92          * @param bool $fulltext
  93          * @return string
  94          */
  95         private function getQuery( $filteredTerm, $fulltext ) {
  96                 return $this->queryLimit( $this->queryMain( $filteredTerm, $fulltext ) . ' ' .
  97                         $this->queryNamespaces() . ' ' .
  98                         $this->queryRanking( $filteredTerm, $fulltext ) . ' ' );
  99         }
 100
 101         /**
 102          * Picks which field to index on, depending on what type of query.
 103          *
 104          * @param bool $fulltext
 105          * @return string
 106          */
 107         function getIndexField( $fulltext ) {
 108                 return $fulltext ? 'si_text' : 'si_title';
 109         }
 110
 111         /**
 112          * Get the base part of the search query.
 113          *
 114          * @param string $filteredTerm
 115          * @param bool $fulltext
 116          * @return string
 117          */
 118         private function queryMain( $filteredTerm, $fulltext ) {
 119                 $match = $this->parseQuery( $filteredTerm, $fulltext );
 120                 $page = $this->db->tableName( 'page' );
 121                 $searchindex = $this->db->tableName( 'searchindex' );
 122
 123                 return 'SELECT page_id, page_namespace, page_title, ftindex.[RANK]' .
 124                         "FROM $page,FREETEXTTABLE($searchindex , $match, LANGUAGE 'English') as ftindex " .
 125                         'WHERE page_id=ftindex.[KEY] ';
 126         }
 127
 128         /** @todo document
 129          * @param string $filteredText
 130          * @param bool $fulltext
 131          * @return string
 132          */
 133         private function parseQuery( $filteredText, $fulltext ) {
 134                 global $wgContLang;
 135                 $lc = $this->legalSearchChars( self::CHARS_NO_SYNTAX );
 136                 $this->searchTerms = [];
 137
 138                 # @todo FIXME: This doesn't handle parenthetical expressions.
 139                 $m = [];
 140                 $q = [];
 141
 142                 if ( preg_match_all( '/([-+<>~]?)(([' . $lc . ']+)(\*?)|"[^"]*")/',
 143                         $filteredText, $m, PREG_SET_ORDER ) ) {
 144                         foreach ( $m as $terms ) {
 145                                 $q[] = $terms[1] . $wgContLang->normalizeForSearch( $terms[2] );
 146
 147                                 if ( !empty( $terms[3] ) ) {
 148                                         $regexp = preg_quote( $terms[3], '/' );
 149                                         if ( $terms[4] ) {
 150                                                 $regexp .= "[0-9A-Za-z_]+";
 151                                         }
 152                                 } else {
 153                                         $regexp = preg_quote( str_replace( '"', '', $terms[2] ), '/' );
 154                                 }
 155                                 $this->searchTerms[] = $regexp;
 156                         }
 157                 }
 158
 159                 $searchon = $this->db->addQuotes( implode( ',', $q ) );
 160                 $field = $this->getIndexField( $fulltext );
 161                 return "$field, $searchon";
 162         }
 163
 164         /**
 165          * Create or update the search index record for the given page.
 166          * Title and text should be pre-processed.
 167          *
 168          * @param int $id
 169          * @param string $title
 170          * @param string $text
 171          * @return bool|ResultWrapper
 172          */
 173         function update( $id, $title, $text ) {
 174                 // We store the column data as UTF-8 byte order marked binary stream
 175                 // because we are invoking the plain text IFilter on it so that, and we want it
 176                 // to properly decode the stream as UTF-8.  SQL doesn't support UTF8 as a data type
 177                 // but the indexer will correctly handle it by this method.  Since all we are doing
 178                 // is passing this data to the indexer and never retrieving it via PHP, this will save space
 179                 $table = $this->db->tableName( 'searchindex' );
 180                 $utf8bom = '0xEFBBBF';
 181                 $si_title = $utf8bom . bin2hex( $title );
 182                 $si_text = $utf8bom . bin2hex( $text );
 183                 $sql = "DELETE FROM $table WHERE si_page = $id;";
 184                 $sql .= "INSERT INTO $table (si_page, si_title, si_text) VALUES ($id, $si_title, $si_text)";
 185                 return $this->db->query( $sql, 'SearchMssql::update' );
 186         }
 187
 188         /**
 189          * Update a search index record's title only.
 190          * Title should be pre-processed.
 191          *
 192          * @param int $id
 193          * @param string $title
 194          * @return bool|ResultWrapper
 195          */
 196         function updateTitle( $id, $title ) {
 197                 $table = $this->db->tableName( 'searchindex' );
 198
 199                 // see update for why we are using the utf8bom
 200                 $utf8bom = '0xEFBBBF';
 201                 $si_title = $utf8bom . bin2hex( $title );
 202                 $sql = "DELETE FROM $table WHERE si_page = $id;";
 203                 $sql .= "INSERT INTO $table (si_page, si_title, si_text) VALUES ($id, $si_title, 0x00)";
 204                 return $this->db->query( $sql, 'SearchMssql::updateTitle' );
 205         }
 206 }