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