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 (' . $namespaces . ')';
 122         }
 123
 124         /**
 125          * Return a LIMIT clause to limit results on the query.
 126          * @return String
 127          */
 128         function queryLimit($sql) {
 129                 return $this->db->limitResult($sql, $this->limit, $this->offset);
 130         }
 131
 132         /**
 133          * Does not do anything for generic search engine
 134          * subclasses may define this though
 135          * @return String
 136          */
 137         function queryRanking($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          */
 147         function getQuery( $filteredTerm, $fulltext ) {
 148                 return $this->queryLimit($this->queryMain($filteredTerm, $fulltext) . ' ' .
 149                         $this->queryRedirect() . ' ' .
 150                         $this->queryNamespaces() . ' ' .
 151                         $this->queryRanking( $filteredTerm, $fulltext ) . ' ');
 152         }
 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 ' . $match;
 178         }
 179
 180         /**
 181          * Parse a user input search string, and return an SQL fragment to be used
 182          * as part of a WHERE clause
 183          */
 184         function parseQuery($filteredText, $fulltext) {
 185                 global $wgContLang;
 186                 $lc = SearchEngine::legalSearchChars();
 187                 $this->searchTerms = array();
 188
 189                 # FIXME: This doesn't handle parenthetical expressions.
 190                 $m = array();
 191                 $searchon = '';
 192                 if (preg_match_all('/([-+<>~]?)(([' . $lc . ']+)(\*?)|"[^"]*")/',
 193                           $filteredText, $m, PREG_SET_ORDER)) {
 194                         foreach($m as $terms) {
 195                                 // Search terms in all variant forms, only
 196                                 // apply on wiki with LanguageConverter
 197                                 $temp_terms = $wgContLang->autoConvertToAllVariants( $terms[2] );
 198                                 if( is_array( $temp_terms )) {
 199                                         $temp_terms = array_unique( array_values( $temp_terms ));
 200                                         foreach( $temp_terms as $t ) {
 201                                                 $searchon .= ($terms[1] == '-' ? ' ~' : ' & ') . $this->escapeTerm( $t );
 202                                         }
 203                                 }
 204                                 else {
 205                                         $searchon .= ($terms[1] == '-' ? ' ~' : ' & ') . $this->escapeTerm( $terms[2] );
 206                                 }
 207                                 if (!empty($terms[3])) {
 208                                         $regexp = preg_quote( $terms[3], '/' );
 209                                         if ($terms[4])
 210                                                 $regexp .= "[0-9A-Za-z_]+";
 211                                 } else {
 212                                         $regexp = preg_quote(str_replace('"', '', $terms[2]), '/');
 213                                 }
 214                                 $this->searchTerms[] = $regexp;
 215                         }
 216                 }
 217
 218
 219                 $searchon = $this->db->addQuotes(ltrim($searchon, ' &'));
 220                 $field = $this->getIndexField($fulltext);
 221                 return " CONTAINS($field, $searchon, 1) > 0 ";
 222         }
 223
 224         private function escapeTerm($t) {
 225                 global $wgContLang;
 226                 $t = $wgContLang->normalizeForSearch($t);
 227                 $t = isset($this->reservedWords[strtoupper($t)]) ? '{'.$t.'}' : $t;
 228                 $t = preg_replace('/^"(.*)"$/', '($1)', $t);
 229                 $t = preg_replace('/([-&|])/', '\\\\$1', $t);
 230                 return $t;
 231         }
 232         /**
 233          * Create or update the search index record for the given page.
 234          * Title and text should be pre-processed.
 235          *
 236          * @param $id Integer
 237          * @param $title String
 238          * @param $text String
 239          */
 240         function update($id, $title, $text) {
 241                 $dbw = wfGetDB(DB_MASTER);
 242                 $dbw->replace('searchindex',
 243                         array('si_page'),
 244                         array(
 245                                 'si_page' => $id,
 246                                 'si_title' => $title,
 247                                 'si_text' => $text
 248                         ), 'SearchOracle::update' );
 249
 250                 // Sync the index
 251                 // We need to specify the DB name (i.e. user/schema) here so that
 252                 // it can work from the installer, where
 253                 //     ALTER SESSION SET CURRENT_SCHEMA = ...
 254                 // was used.
 255                 $dbw->query( "CALL ctx_ddl.sync_index(" .
 256                         $dbw->addQuotes( $dbw->getDBname() . '.si_text_idx' ) . ")" );
 257                 $dbw->query( "CALL ctx_ddl.sync_index(" .
 258                         $dbw->addQuotes( $dbw->getDBname() . '.si_title_idx' ) . ")" );
 259         }
 260
 261         /**
 262          * Update a search index record's title only.
 263          * Title should be pre-processed.
 264          *
 265          * @param $id Integer
 266          * @param $title String
 267          */
 268         function updateTitle($id, $title) {
 269                 $dbw = wfGetDB(DB_MASTER);
 270
 271                 $dbw->update('searchindex',
 272                         array('si_title' => $title),
 273                         array('si_page'  => $id),
 274                         'SearchOracle::updateTitle',
 275                         array());
 276         }
 277
 278
 279         public static function legalSearchChars() {
 280                 return "\"" . parent::legalSearchChars();
 281         }
 282 }