includes/search/SearchOracle.php

   1 <?php
   2 /**
   3  * Oracle search engine
   4  *
   5  * Copyright © 2004 Brion Vibber <brion@pobox.com>
   6  * https://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 SearchDatabase {
  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          * Perform a full text search query and return a result set.
  64          *
  65          * @param string $term raw search term
  66          * @return SqlSearchResultSet
  67          */
  68         function searchText( $term ) {
  69                 if ( $term == '' ) {
  70                         return new SqlSearchResultSet( false, '' );
  71                 }
  72
  73                 $resultSet = $this->db->resultObject( $this->db->query( $this->getQuery( $this->filter( $term ), true ) ) );
  74                 return new SqlSearchResultSet( $resultSet, $this->searchTerms );
  75         }
  76
  77         /**
  78          * Perform a title-only search query and return a result set.
  79          *
  80          * @param string $term raw search term
  81          * @return SqlSearchResultSet
  82          */
  83         function searchTitle( $term ) {
  84                 if ( $term == '' ) {
  85                         return new SqlSearchResultSet( false, '' );
  86                 }
  87
  88                 $resultSet = $this->db->resultObject( $this->db->query( $this->getQuery( $this->filter( $term ), false ) ) );
  89                 return new MySQLSearchResultSet( $resultSet, $this->searchTerms );
  90         }
  91
  92         /**
  93          * Return a partial WHERE clause to exclude redirects, if so set
  94          * @return String
  95          */
  96         function queryRedirect() {
  97                 if ( $this->showRedirects ) {
  98                         return '';
  99                 } else {
 100                         return 'AND page_is_redirect=0';
 101                 }
 102         }
 103
 104         /**
 105          * Return a partial WHERE clause to limit the search to the given namespaces
 106          * @return String
 107          */
 108         function queryNamespaces() {
 109                 if ( is_null( $this->namespaces ) ) {
 110                         return '';
 111                 }
 112                 if ( !count( $this->namespaces ) ) {
 113                         $namespaces = '0';
 114                 } else {
 115                         $namespaces = $this->db->makeList( $this->namespaces );
 116                 }
 117                 return 'AND page_namespace IN (' . $namespaces . ')';
 118         }
 119
 120         /**
 121          * Return a LIMIT clause to limit results on the query.
 122          *
 123          * @param $sql string
 124          *
 125          * @return String
 126          */
 127         function queryLimit( $sql ) {
 128                 return $this->db->limitResult( $sql, $this->limit, $this->offset );
 129         }
 130
 131         /**
 132          * Does not do anything for generic search engine
 133          * subclasses may define this though
 134          *
 135          * @return String
 136          */
 137         function queryRanking( $filteredTerm, $fulltext ) {
 138                 return ' ORDER BY score(1)';
 139         }
 140
 141         /**
 142          * Construct the full SQL query to do the search.
 143          * The guts shoulds be constructed in queryMain()
 144          * @param $filteredTerm String
 145          * @param $fulltext Boolean
 146          * @return String
 147          */
 148         function getQuery( $filteredTerm, $fulltext ) {
 149                 return $this->queryLimit( $this->queryMain( $filteredTerm, $fulltext ) . ' ' .
 150                         $this->queryRedirect() . ' ' .
 151                         $this->queryNamespaces() . ' ' .
 152                         $this->queryRanking( $filteredTerm, $fulltext ) . ' ' );
 153         }
 154
 155         /**
 156          * Picks which field to index on, depending on what type of query.
 157          * @param $fulltext Boolean
 158          * @return String
 159          */
 160         function getIndexField( $fulltext ) {
 161                 return $fulltext ? 'si_text' : 'si_title';
 162         }
 163
 164         /**
 165          * Get the base part of the search query.
 166          *
 167          * @param $filteredTerm String
 168          * @param $fulltext Boolean
 169          * @return String
 170          */
 171         function queryMain( $filteredTerm, $fulltext ) {
 172                 $match = $this->parseQuery( $filteredTerm, $fulltext );
 173                 $page = $this->db->tableName( 'page' );
 174                 $searchindex = $this->db->tableName( 'searchindex' );
 175                 return 'SELECT page_id, page_namespace, page_title ' .
 176                         "FROM $page,$searchindex " .
 177                         'WHERE page_id=si_page AND ' . $match;
 178         }
 179
 180         /**
 181          * Parse a user input search string, and return an SQL fragment to be used
 182          * as part of a WHERE clause
 183          * @return string
 184          */
 185         function parseQuery( $filteredText, $fulltext ) {
 186                 global $wgContLang;
 187                 $lc = SearchEngine::legalSearchChars();
 188                 $this->searchTerms = array();
 189
 190                 # @todo FIXME: This doesn't handle parenthetical expressions.
 191                 $m = array();
 192                 $searchon = '';
 193                 if ( preg_match_all( '/([-+<>~]?)(([' . $lc . ']+)(\*?)|"[^"]*")/',
 194                                 $filteredText, $m, PREG_SET_ORDER ) ) {
 195                         foreach ( $m as $terms ) {
 196                                 // Search terms in all variant forms, only
 197                                 // apply on wiki with LanguageConverter
 198                                 $temp_terms = $wgContLang->autoConvertToAllVariants( $terms[2] );
 199                                 if ( is_array( $temp_terms ) ) {
 200                                         $temp_terms = array_unique( array_values( $temp_terms ) );
 201                                         foreach ( $temp_terms as $t ) {
 202                                                 $searchon .= ( $terms[1] == '-' ? ' ~' : ' & ' ) . $this->escapeTerm( $t );
 203                                         }
 204                                 }
 205                                 else {
 206                                         $searchon .= ( $terms[1] == '-' ? ' ~' : ' & ' ) . $this->escapeTerm( $terms[2] );
 207                                 }
 208                                 if ( !empty( $terms[3] ) ) {
 209                                         $regexp = preg_quote( $terms[3], '/' );
 210                                         if ( $terms[4] ) {
 211                                                 $regexp .= "[0-9A-Za-z_]+";
 212                                         }
 213                                 } else {
 214                                         $regexp = preg_quote( str_replace( '"', '', $terms[2] ), '/' );
 215                                 }
 216                                 $this->searchTerms[] = $regexp;
 217                         }
 218                 }
 219
 220                 $searchon = $this->db->addQuotes( ltrim( $searchon, ' &' ) );
 221                 $field = $this->getIndexField( $fulltext );
 222                 return " CONTAINS($field, $searchon, 1) > 0 ";
 223         }
 224
 225         private function escapeTerm( $t ) {
 226                 global $wgContLang;
 227                 $t = $wgContLang->normalizeForSearch( $t );
 228                 $t = isset( $this->reservedWords[strtoupper( $t )] ) ? '{' . $t . '}' : $t;
 229                 $t = preg_replace( '/^"(.*)"$/', '($1)', $t );
 230                 $t = preg_replace( '/([-&|])/', '\\\\$1', $t );
 231                 return $t;
 232         }
 233
 234         /**
 235          * Create or update the search index record for the given page.
 236          * Title and text should be pre-processed.
 237          *
 238          * @param $id Integer
 239          * @param $title String
 240          * @param $text String
 241          */
 242         function update( $id, $title, $text ) {
 243                 $dbw = wfGetDB( DB_MASTER );
 244                 $dbw->replace( 'searchindex',
 245                         array( 'si_page' ),
 246                         array(
 247                                 'si_page' => $id,
 248                                 'si_title' => $title,
 249                                 'si_text' => $text
 250                         ), 'SearchOracle::update' );
 251
 252                 // Sync the index
 253                 // We need to specify the DB name (i.e. user/schema) here so that
 254                 // it can work from the installer, where
 255                 //     ALTER SESSION SET CURRENT_SCHEMA = ...
 256                 // was used.
 257                 $dbw->query( "CALL ctx_ddl.sync_index(" .
 258                         $dbw->addQuotes( $dbw->getDBname() . '.' . $dbw->tableName( 'si_text_idx', 'raw' ) ) . ")" );
 259                 $dbw->query( "CALL ctx_ddl.sync_index(" .
 260                         $dbw->addQuotes( $dbw->getDBname() . '.' . $dbw->tableName( 'si_title_idx', 'raw' ) ) . ")" );
 261         }
 262
 263         /**
 264          * Update a search index record's title only.
 265          * Title should be pre-processed.
 266          *
 267          * @param $id Integer
 268          * @param $title String
 269          */
 270         function updateTitle( $id, $title ) {
 271                 $dbw = wfGetDB( DB_MASTER );
 272
 273                 $dbw->update( 'searchindex',
 274                         array( 'si_title' => $title ),
 275                         array( 'si_page' => $id ),
 276                         'SearchOracle::updateTitle',
 277                         array() );
 278         }
 279
 280         public static function legalSearchChars() {
 281                 return "\"" . parent::legalSearchChars();
 282         }
 283 }