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