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         private $reservedWords = array(
  33                 '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         /**
  62          * Perform a full text search query and return a result set.
  63          *
  64          * @param string $term Raw search term
  65          * @return SqlSearchResultSet
  66          */
  67         function searchText( $term ) {
  68                 if ( $term == '' ) {
  69                         return new SqlSearchResultSet( false, '' );
  70                 }
  71
  72                 $resultSet = $this->db->resultObject(
  73                         $this->db->query( $this->getQuery( $this->filter( $term ), true ) )
  74                 );
  75
  76                 return new SqlSearchResultSet( $resultSet, $this->searchTerms );
  77         }
  78
  79         /**
  80          * Perform a title-only search query and return a result set.
  81          *
  82          * @param string $term Raw search term
  83          * @return SqlSearchResultSet
  84          */
  85         function searchTitle( $term ) {
  86                 if ( $term == '' ) {
  87                         return new SqlSearchResultSet( false, '' );
  88                 }
  89
  90                 $resultSet = $this->db->resultObject(
  91                         $this->db->query( $this->getQuery( $this->filter( $term ), false ) )
  92                 );
  93
  94                 return new SqlSearchResultSet( $resultSet, $this->searchTerms );
  95         }
  96
  97         /**
  98          * Return a partial WHERE clause to limit the search to the given namespaces
  99          * @return string
 100          */
 101         function queryNamespaces() {
 102                 if ( is_null( $this->namespaces ) ) {
 103                         return '';
 104                 }
 105                 if ( !count( $this->namespaces ) ) {
 106                         $namespaces = '0';
 107                 } else {
 108                         $namespaces = $this->db->makeList( $this->namespaces );
 109                 }
 110                 return 'AND page_namespace IN (' . $namespaces . ')';
 111         }
 112
 113         /**
 114          * Return a LIMIT clause to limit results on the query.
 115          *
 116          * @param string $sql
 117          *
 118          * @return string
 119          */
 120         function queryLimit( $sql ) {
 121                 return $this->db->limitResult( $sql, $this->limit, $this->offset );
 122         }
 123
 124         /**
 125          * Does not do anything for generic search engine
 126          * subclasses may define this though
 127          *
 128          * @return string
 129          */
 130         function queryRanking( $filteredTerm, $fulltext ) {
 131                 return ' ORDER BY score(1)';
 132         }
 133
 134         /**
 135          * Construct the full SQL query to do the search.
 136          * The guts shoulds be constructed in queryMain()
 137          * @param string $filteredTerm
 138          * @param bool $fulltext
 139          * @return string
 140          */
 141         function getQuery( $filteredTerm, $fulltext ) {
 142                 return $this->queryLimit( $this->queryMain( $filteredTerm, $fulltext ) . ' ' .
 143                         $this->queryNamespaces() . ' ' .
 144                         $this->queryRanking( $filteredTerm, $fulltext ) . ' ' );
 145         }
 146
 147         /**
 148          * Picks which field to index on, depending on what type of query.
 149          * @param bool $fulltext
 150          * @return string
 151          */
 152         function getIndexField( $fulltext ) {
 153                 return $fulltext ? 'si_text' : 'si_title';
 154         }
 155
 156         /**
 157          * Get the base part of the search query.
 158          *
 159          * @param string $filteredTerm
 160          * @param bool $fulltext
 161          * @return string
 162          */
 163         function queryMain( $filteredTerm, $fulltext ) {
 164                 $match = $this->parseQuery( $filteredTerm, $fulltext );
 165                 $page = $this->db->tableName( 'page' );
 166                 $searchindex = $this->db->tableName( 'searchindex' );
 167                 return 'SELECT page_id, page_namespace, page_title ' .
 168                         "FROM $page,$searchindex " .
 169                         'WHERE page_id=si_page AND ' . $match;
 170         }
 171
 172         /**
 173          * Parse a user input search string, and return an SQL fragment to be used
 174          * as part of a WHERE clause
 175          * @return string
 176          */
 177         function parseQuery( $filteredText, $fulltext ) {
 178                 global $wgContLang;
 179                 $lc = $this->legalSearchChars();
 180                 $this->searchTerms = array();
 181
 182                 # @todo FIXME: This doesn't handle parenthetical expressions.
 183                 $m = array();
 184                 $searchon = '';
 185                 if ( preg_match_all( '/([-+<>~]?)(([' . $lc . ']+)(\*?)|"[^"]*")/',
 186                                 $filteredText, $m, PREG_SET_ORDER ) ) {
 187                         foreach ( $m as $terms ) {
 188                                 // Search terms in all variant forms, only
 189                                 // apply on wiki with LanguageConverter
 190                                 $temp_terms = $wgContLang->autoConvertToAllVariants( $terms[2] );
 191                                 if ( is_array( $temp_terms ) ) {
 192                                         $temp_terms = array_unique( array_values( $temp_terms ) );
 193                                         foreach ( $temp_terms as $t ) {
 194                                                 $searchon .= ( $terms[1] == '-' ? ' ~' : ' & ' ) . $this->escapeTerm( $t );
 195                                         }
 196                                 }
 197                                 else {
 198                                         $searchon .= ( $terms[1] == '-' ? ' ~' : ' & ' ) . $this->escapeTerm( $terms[2] );
 199                                 }
 200                                 if ( !empty( $terms[3] ) ) {
 201                                         $regexp = preg_quote( $terms[3], '/' );
 202                                         if ( $terms[4] ) {
 203                                                 $regexp .= "[0-9A-Za-z_]+";
 204                                         }
 205                                 } else {
 206                                         $regexp = preg_quote( str_replace( '"', '', $terms[2] ), '/' );
 207                                 }
 208                                 $this->searchTerms[] = $regexp;
 209                         }
 210                 }
 211
 212                 $searchon = $this->db->addQuotes( ltrim( $searchon, ' &' ) );
 213                 $field = $this->getIndexField( $fulltext );
 214                 return " CONTAINS($field, $searchon, 1) > 0 ";
 215         }
 216
 217         private function escapeTerm( $t ) {
 218                 global $wgContLang;
 219                 $t = $wgContLang->normalizeForSearch( $t );
 220                 $t = isset( $this->reservedWords[strtoupper( $t )] ) ? '{' . $t . '}' : $t;
 221                 $t = preg_replace( '/^"(.*)"$/', '($1)', $t );
 222                 $t = preg_replace( '/([-&|])/', '\\\\$1', $t );
 223                 return $t;
 224         }
 225
 226         /**
 227          * Create or update the search index record for the given page.
 228          * Title and text should be pre-processed.
 229          *
 230          * @param int $id
 231          * @param string $title
 232          * @param string $text
 233          */
 234         function update( $id, $title, $text ) {
 235                 $dbw = wfGetDB( DB_MASTER );
 236                 $dbw->replace( 'searchindex',
 237                         array( 'si_page' ),
 238                         array(
 239                                 'si_page' => $id,
 240                                 'si_title' => $title,
 241                                 'si_text' => $text
 242                         ), 'SearchOracle::update' );
 243
 244                 // Sync the index
 245                 // We need to specify the DB name (i.e. user/schema) here so that
 246                 // it can work from the installer, where
 247                 //     ALTER SESSION SET CURRENT_SCHEMA = ...
 248                 // was used.
 249                 $dbw->query( "CALL ctx_ddl.sync_index(" .
 250                         $dbw->addQuotes( $dbw->getDBname() . '.' . $dbw->tableName( 'si_text_idx', 'raw' ) ) . ")" );
 251                 $dbw->query( "CALL ctx_ddl.sync_index(" .
 252                         $dbw->addQuotes( $dbw->getDBname() . '.' . $dbw->tableName( 'si_title_idx', 'raw' ) ) . ")" );
 253         }
 254
 255         /**
 256          * Update a search index record's title only.
 257          * Title should be pre-processed.
 258          *
 259          * @param int $id
 260          * @param string $title
 261          */
 262         function updateTitle( $id, $title ) {
 263                 $dbw = wfGetDB( DB_MASTER );
 264
 265                 $dbw->update( 'searchindex',
 266                         array( 'si_title' => $title ),
 267                         array( 'si_page' => $id ),
 268                         'SearchOracle::updateTitle',
 269                         array() );
 270         }
 271
 272         public static function legalSearchChars() {
 273                 return "\"" . parent::legalSearchChars();
 274         }
 275 }