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 $term String: 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 $term String: 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         /**
  98          * Return a partial WHERE clause to exclude redirects, if so set
  99          * @return String
 100          */
 101         function queryRedirect() {
 102                 if ( $this->showRedirects ) {
 103                         return '';
 104                 } else {
 105                         return 'AND page_is_redirect=0';
 106                 }
 107         }
 108
 109         /**
 110          * Return a partial WHERE clause to limit the search to the given namespaces
 111          * @return String
 112          */
 113         function queryNamespaces() {
 114                 if( is_null( $this->namespaces ) )
 115                         return '';
 116                 if ( !count( $this->namespaces ) ) {
 117                         $namespaces = '0';
 118                 } else {
 119                         $namespaces = $this->db->makeList( $this->namespaces );
 120                 }
 121                 return 'AND page_namespace IN (' . $namespaces . ')';
 122         }
 123
 124         /**
 125          * Return a LIMIT clause to limit results on the query.
 126          *
 127          * @param $sql string
 128          *
 129          * @return String
 130          */
 131         function queryLimit( $sql ) {
 132                 return $this->db->limitResult( $sql, $this->limit, $this->offset );
 133         }
 134
 135         /**
 136          * Does not do anything for generic search engine
 137          * subclasses may define this though
 138          *
 139          * @return String
 140          */
 141         function queryRanking( $filteredTerm, $fulltext ) {
 142                 return ' ORDER BY score(1)';
 143         }
 144
 145         /**
 146          * Construct the full SQL query to do the search.
 147          * The guts shoulds be constructed in queryMain()
 148          * @param $filteredTerm String
 149          * @param $fulltext Boolean
 150          * @return String
 151          */
 152         function getQuery( $filteredTerm, $fulltext ) {
 153                 return $this->queryLimit( $this->queryMain( $filteredTerm, $fulltext ) . ' ' .
 154                         $this->queryRedirect() . ' ' .
 155                         $this->queryNamespaces() . ' ' .
 156                         $this->queryRanking( $filteredTerm, $fulltext ) . ' ' );
 157         }
 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
 225                 $searchon = $this->db->addQuotes( ltrim( $searchon, ' &' ) );
 226                 $field = $this->getIndexField( $fulltext );
 227                 return " CONTAINS($field, $searchon, 1) > 0 ";
 228         }
 229
 230         private function escapeTerm( $t ) {
 231                 global $wgContLang;
 232                 $t = $wgContLang->normalizeForSearch( $t );
 233                 $t = isset( $this->reservedWords[strtoupper( $t )] ) ? '{'.$t.'}' : $t;
 234                 $t = preg_replace('/^"(.*)"$/', '($1)', $t);
 235                 $t = preg_replace('/([-&|])/', '\\\\$1', $t);
 236                 return $t;
 237         }
 238         /**
 239          * Create or update the search index record for the given page.
 240          * Title and text should be pre-processed.
 241          *
 242          * @param $id Integer
 243          * @param $title String
 244          * @param $text String
 245          */
 246         function update( $id, $title, $text ) {
 247                 $dbw = wfGetDB( DB_MASTER );
 248                 $dbw->replace( 'searchindex',
 249                         array( 'si_page' ),
 250                         array(
 251                                 'si_page' => $id,
 252                                 'si_title' => $title,
 253                                 'si_text' => $text
 254                         ), 'SearchOracle::update' );
 255
 256                 // Sync the index
 257                 // We need to specify the DB name (i.e. user/schema) here so that
 258                 // it can work from the installer, where
 259                 //     ALTER SESSION SET CURRENT_SCHEMA = ...
 260                 // was used.
 261                 $dbw->query( "CALL ctx_ddl.sync_index(" .
 262                         $dbw->addQuotes( $dbw->getDBname() . '.' . $dbw->tableName( 'si_text_idx', 'raw' ) ) . ")" );
 263                 $dbw->query( "CALL ctx_ddl.sync_index(" .
 264                         $dbw->addQuotes( $dbw->getDBname() . '.' . $dbw->tableName( 'si_title_idx', 'raw' ) ) . ")" );
 265         }
 266
 267         /**
 268          * Update a search index record's title only.
 269          * Title should be pre-processed.
 270          *
 271          * @param $id Integer
 272          * @param $title String
 273          */
 274         function updateTitle( $id, $title ) {
 275                 $dbw = wfGetDB( DB_MASTER );
 276
 277                 $dbw->update( 'searchindex',
 278                         array( 'si_title' => $title ),
 279                         array( 'si_page'  => $id ),
 280                         'SearchOracle::updateTitle',
 281                         array() );
 282         }
 283
 284
 285         public static function legalSearchChars() {
 286                 return "\"" . parent::legalSearchChars();
 287         }
 288 }