includes/search/SearchOracle.php

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