includes/search/SearchOracle.php

   1 <?php
   2 /**
   3  * Oracle search engine
   4  *
   5  * Copyright © 2004 Brion Vibber <brion@pobox.com>
   6  * http://www.mediawiki.org/
   7  *
   8  * This program is free software; you can redistribute it and/or modify
   9  * it under the terms of the GNU General Public License as published by
  10  * the Free Software Foundation; either version 2 of the License, or
  11  * (at your option) any later version.
  12  *
  13  * This program is distributed in the hope that it will be useful,
  14  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  15  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  16  * GNU General Public License for more details.
  17  *
  18  * You should have received a copy of the GNU General Public License along
  19  * with this program; if not, write to the Free Software Foundation, Inc.,
  20  * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
  21  * http://www.gnu.org/copyleft/gpl.html
  22  *
  23  * @file
  24  * @ingroup Search
  25  */
  26
  27 /**
  28  * Search engine hook base class for Oracle (ConText).
  29  * @ingroup Search
  30  */
  31 class SearchOracle extends SearchEngine {
  32
  33         private $reservedWords = array(
  34                 'ABOUT' => 1,
  35                 'ACCUM' => 1,
  36                 'AND' => 1,
  37                 'BT' => 1,
  38                 'BTG' => 1,
  39                 'BTI' => 1,
  40                 'BTP' => 1,
  41                 'FUZZY' => 1,
  42                 'HASPATH' => 1,
  43                 'INPATH' => 1,
  44                 'MINUS' => 1,
  45                 'NEAR' => 1,
  46                 'NOT' => 1,
  47                 'NT' => 1,
  48                 'NTG' => 1,
  49                 'NTI' => 1,
  50                 'NTP' => 1,
  51                 'OR' => 1,
  52                 'PT' => 1,
  53                 'RT' => 1,
  54                 'SQE' => 1,
  55                 'SYN' => 1,
  56                 'TR' => 1,
  57                 'TRSYN' => 1,
  58                 'TT' => 1,
  59                 'WITHIN' => 1,
  60         );
  61
  62         /**
  63          * Creates an instance of this class
  64          * @param $db DatabasePostgres: database object
  65          */
  66         function __construct( $db ) {
  67                 parent::__construct( $db );
  68         }
  69
  70         /**
  71          * Perform a full text search query and return a result set.
  72          *
  73          * @param string $term raw search term
  74          * @return SqlSearchResultSet
  75          */
  76         function searchText( $term ) {
  77                 if ( $term == '' )
  78                         return new SqlSearchResultSet( false, '' );
  79
  80                 $resultSet = $this->db->resultObject( $this->db->query( $this->getQuery( $this->filter( $term ), true ) ) );
  81                 return new SqlSearchResultSet( $resultSet, $this->searchTerms );
  82         }
  83
  84         /**
  85          * Perform a title-only search query and return a result set.
  86          *
  87          * @param string $term raw search term
  88          * @return SqlSearchResultSet
  89          */
  90         function searchTitle( $term ) {
  91                 if ( $term == '' )
  92                         return new SqlSearchResultSet( false, '' );
  93
  94                 $resultSet = $this->db->resultObject( $this->db->query( $this->getQuery( $this->filter( $term ), false ) ) );
  95                 return new MySQLSearchResultSet( $resultSet, $this->searchTerms );
  96         }
  97
  98         /**
  99          * Return a partial WHERE clause to exclude redirects, if so set
 100          * @return String
 101          */
 102         function queryRedirect() {
 103                 if ( $this->showRedirects ) {
 104                         return '';
 105                 } else {
 106                         return 'AND page_is_redirect=0';
 107                 }
 108         }
 109
 110         /**
 111          * Return a partial WHERE clause to limit the search to the given namespaces
 112          * @return String
 113          */
 114         function queryNamespaces() {
 115                 if( is_null( $this->namespaces ) )
 116                         return '';
 117                 if ( !count( $this->namespaces ) ) {
 118                         $namespaces = '0';
 119                 } else {
 120                         $namespaces = $this->db->makeList( $this->namespaces );
 121                 }
 122                 return 'AND page_namespace IN (' . $namespaces . ')';
 123         }
 124
 125         /**
 126          * Return a LIMIT clause to limit results on the query.
 127          *
 128          * @param $sql string
 129          *
 130          * @return String
 131          */
 132         function queryLimit( $sql ) {
 133                 return $this->db->limitResult( $sql, $this->limit, $this->offset );
 134         }
 135
 136         /**
 137          * Does not do anything for generic search engine
 138          * subclasses may define this though
 139          *
 140          * @return String
 141          */
 142         function queryRanking( $filteredTerm, $fulltext ) {
 143                 return ' ORDER BY score(1)';
 144         }
 145
 146         /**
 147          * Construct the full SQL query to do the search.
 148          * The guts shoulds be constructed in queryMain()
 149          * @param $filteredTerm String
 150          * @param $fulltext Boolean
 151          * @return String
 152          */
 153         function getQuery( $filteredTerm, $fulltext ) {
 154                 return $this->queryLimit( $this->queryMain( $filteredTerm, $fulltext ) . ' ' .
 155                         $this->queryRedirect() . ' ' .
 156                         $this->queryNamespaces() . ' ' .
 157                         $this->queryRanking( $filteredTerm, $fulltext ) . ' ' );
 158         }
 159
 160         /**
 161          * Picks which field to index on, depending on what type of query.
 162          * @param $fulltext Boolean
 163          * @return String
 164          */
 165         function getIndexField( $fulltext ) {
 166                 return $fulltext ? 'si_text' : 'si_title';
 167         }
 168
 169         /**
 170          * Get the base part of the search query.
 171          *
 172          * @param $filteredTerm String
 173          * @param $fulltext Boolean
 174          * @return String
 175          */
 176         function queryMain( $filteredTerm, $fulltext ) {
 177                 $match = $this->parseQuery( $filteredTerm, $fulltext );
 178                 $page = $this->db->tableName( 'page' );
 179                 $searchindex = $this->db->tableName( 'searchindex' );
 180                 return 'SELECT page_id, page_namespace, page_title ' .
 181                         "FROM $page,$searchindex " .
 182                         'WHERE page_id=si_page AND ' . $match;
 183         }
 184
 185         /**
 186          * Parse a user input search string, and return an SQL fragment to be used
 187          * as part of a WHERE clause
 188          * @return string
 189          */
 190         function parseQuery( $filteredText, $fulltext ) {
 191                 global $wgContLang;
 192                 $lc = SearchEngine::legalSearchChars();
 193                 $this->searchTerms = array();
 194
 195                 # @todo FIXME: This doesn't handle parenthetical expressions.
 196                 $m = array();
 197                 $searchon = '';
 198                 if ( preg_match_all( '/([-+<>~]?)(([' . $lc . ']+)(\*?)|"[^"]*")/',
 199                                 $filteredText, $m, PREG_SET_ORDER ) ) {
 200                         foreach( $m as $terms ) {
 201                                 // Search terms in all variant forms, only
 202                                 // apply on wiki with LanguageConverter
 203                                 $temp_terms = $wgContLang->autoConvertToAllVariants( $terms[2] );
 204                                 if( is_array( $temp_terms ) ) {
 205                                         $temp_terms = array_unique( array_values( $temp_terms ) );
 206                                         foreach( $temp_terms as $t ) {
 207                                                 $searchon .= ($terms[1] == '-' ? ' ~' : ' & ') . $this->escapeTerm( $t );
 208                                         }
 209                                 }
 210                                 else {
 211                                         $searchon .= ($terms[1] == '-' ? ' ~' : ' & ') . $this->escapeTerm( $terms[2] );
 212                                 }
 213                                 if ( !empty( $terms[3] ) ) {
 214                                         $regexp = preg_quote( $terms[3], '/' );
 215                                         if ( $terms[4] )
 216                                                 $regexp .= "[0-9A-Za-z_]+";
 217                                 } else {
 218                                         $regexp = preg_quote( str_replace( '"', '', $terms[2] ), '/' );
 219                                 }
 220                                 $this->searchTerms[] = $regexp;
 221                         }
 222                 }
 223
 224                 $searchon = $this->db->addQuotes( ltrim( $searchon, ' &' ) );
 225                 $field = $this->getIndexField( $fulltext );
 226                 return " CONTAINS($field, $searchon, 1) > 0 ";
 227         }
 228
 229         private function escapeTerm( $t ) {
 230                 global $wgContLang;
 231                 $t = $wgContLang->normalizeForSearch( $t );
 232                 $t = isset( $this->reservedWords[strtoupper( $t )] ) ? '{' . $t . '}' : $t;
 233                 $t = preg_replace( '/^"(.*)"$/', '($1)', $t );
 234                 $t = preg_replace( '/([-&|])/', '\\\\$1', $t );
 235                 return $t;
 236         }
 237         /**
 238          * Create or update the search index record for the given page.
 239          * Title and text should be pre-processed.
 240          *
 241          * @param $id Integer
 242          * @param $title String
 243          * @param $text String
 244          */
 245         function update( $id, $title, $text ) {
 246                 $dbw = wfGetDB( DB_MASTER );
 247                 $dbw->replace( 'searchindex',
 248                         array( 'si_page' ),
 249                         array(
 250                                 'si_page' => $id,
 251                                 'si_title' => $title,
 252                                 'si_text' => $text
 253                         ), 'SearchOracle::update' );
 254
 255                 // Sync the index
 256                 // We need to specify the DB name (i.e. user/schema) here so that
 257                 // it can work from the installer, where
 258                 //     ALTER SESSION SET CURRENT_SCHEMA = ...
 259                 // was used.
 260                 $dbw->query( "CALL ctx_ddl.sync_index(" .
 261                         $dbw->addQuotes( $dbw->getDBname() . '.' . $dbw->tableName( 'si_text_idx', 'raw' ) ) . ")" );
 262                 $dbw->query( "CALL ctx_ddl.sync_index(" .
 263                         $dbw->addQuotes( $dbw->getDBname() . '.' . $dbw->tableName( 'si_title_idx', 'raw' ) ) . ")" );
 264         }
 265
 266         /**
 267          * Update a search index record's title only.
 268          * Title should be pre-processed.
 269          *
 270          * @param $id Integer
 271          * @param $title String
 272          */
 273         function updateTitle( $id, $title ) {
 274                 $dbw = wfGetDB( DB_MASTER );
 275
 276                 $dbw->update( 'searchindex',
 277                         array( 'si_title' => $title ),
 278                         array( 'si_page'  => $id ),
 279                         'SearchOracle::updateTitle',
 280                         array() );
 281         }
 282
 283         public static function legalSearchChars() {
 284                 return "\"" . parent::legalSearchChars();
 285         }
 286 }