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          * @param string $filteredTerm
 129          * @param bool $fulltext
 130          * @return string
 131          */
 132         function queryRanking( $filteredTerm, $fulltext ) {
 133                 return ' ORDER BY score(1)';
 134         }
 135
 136         /**
 137          * Construct the full SQL query to do the search.
 138          * The guts shoulds be constructed in queryMain()
 139          * @param string $filteredTerm
 140          * @param bool $fulltext
 141          * @return string
 142          */
 143         function getQuery( $filteredTerm, $fulltext ) {
 144                 return $this->queryLimit( $this->queryMain( $filteredTerm, $fulltext ) . ' ' .
 145                         $this->queryNamespaces() . ' ' .
 146                         $this->queryRanking( $filteredTerm, $fulltext ) . ' ' );
 147         }
 148
 149         /**
 150          * Picks which field to index on, depending on what type of query.
 151          * @param bool $fulltext
 152          * @return string
 153          */
 154         function getIndexField( $fulltext ) {
 155                 return $fulltext ? 'si_text' : 'si_title';
 156         }
 157
 158         /**
 159          * Get the base part of the search query.
 160          *
 161          * @param string $filteredTerm
 162          * @param bool $fulltext
 163          * @return string
 164          */
 165         function queryMain( $filteredTerm, $fulltext ) {
 166                 $match = $this->parseQuery( $filteredTerm, $fulltext );
 167                 $page = $this->db->tableName( 'page' );
 168                 $searchindex = $this->db->tableName( 'searchindex' );
 169                 return 'SELECT page_id, page_namespace, page_title ' .
 170                         "FROM $page,$searchindex " .
 171                         'WHERE page_id=si_page AND ' . $match;
 172         }
 173
 174         /**
 175          * Parse a user input search string, and return an SQL fragment to be used
 176          * as part of a WHERE clause
 177          * @param string $filteredText
 178          * @param bool $fulltext
 179          * @return string
 180          */
 181         function parseQuery( $filteredText, $fulltext ) {
 182                 global $wgContLang;
 183                 $lc = $this->legalSearchChars();
 184                 $this->searchTerms = array();
 185
 186                 # @todo FIXME: This doesn't handle parenthetical expressions.
 187                 $m = array();
 188                 $searchon = '';
 189                 if ( preg_match_all( '/([-+<>~]?)(([' . $lc . ']+)(\*?)|"[^"]*")/',
 190                                 $filteredText, $m, PREG_SET_ORDER ) ) {
 191                         foreach ( $m as $terms ) {
 192                                 // Search terms in all variant forms, only
 193                                 // apply on wiki with LanguageConverter
 194                                 $temp_terms = $wgContLang->autoConvertToAllVariants( $terms[2] );
 195                                 if ( is_array( $temp_terms ) ) {
 196                                         $temp_terms = array_unique( array_values( $temp_terms ) );
 197                                         foreach ( $temp_terms as $t ) {
 198                                                 $searchon .= ( $terms[1] == '-' ? ' ~' : ' & ' ) . $this->escapeTerm( $t );
 199                                         }
 200                                 }
 201                                 else {
 202                                         $searchon .= ( $terms[1] == '-' ? ' ~' : ' & ' ) . $this->escapeTerm( $terms[2] );
 203                                 }
 204                                 if ( !empty( $terms[3] ) ) {
 205                                         $regexp = preg_quote( $terms[3], '/' );
 206                                         if ( $terms[4] ) {
 207                                                 $regexp .= "[0-9A-Za-z_]+";
 208                                         }
 209                                 } else {
 210                                         $regexp = preg_quote( str_replace( '"', '', $terms[2] ), '/' );
 211                                 }
 212                                 $this->searchTerms[] = $regexp;
 213                         }
 214                 }
 215
 216                 $searchon = $this->db->addQuotes( ltrim( $searchon, ' &' ) );
 217                 $field = $this->getIndexField( $fulltext );
 218                 return " CONTAINS($field, $searchon, 1) > 0 ";
 219         }
 220
 221         private function escapeTerm( $t ) {
 222                 global $wgContLang;
 223                 $t = $wgContLang->normalizeForSearch( $t );
 224                 $t = isset( $this->reservedWords[strtoupper( $t )] ) ? '{' . $t . '}' : $t;
 225                 $t = preg_replace( '/^"(.*)"$/', '($1)', $t );
 226                 $t = preg_replace( '/([-&|])/', '\\\\$1', $t );
 227                 return $t;
 228         }
 229
 230         /**
 231          * Create or update the search index record for the given page.
 232          * Title and text should be pre-processed.
 233          *
 234          * @param int $id
 235          * @param string $title
 236          * @param string $text
 237          */
 238         function update( $id, $title, $text ) {
 239                 $dbw = wfGetDB( DB_MASTER );
 240                 $dbw->replace( 'searchindex',
 241                         array( 'si_page' ),
 242                         array(
 243                                 'si_page' => $id,
 244                                 'si_title' => $title,
 245                                 'si_text' => $text
 246                         ), 'SearchOracle::update' );
 247
 248                 // Sync the index
 249                 // We need to specify the DB name (i.e. user/schema) here so that
 250                 // it can work from the installer, where
 251                 //     ALTER SESSION SET CURRENT_SCHEMA = ...
 252                 // was used.
 253                 $dbw->query( "CALL ctx_ddl.sync_index(" .
 254                         $dbw->addQuotes( $dbw->getDBname() . '.' . $dbw->tableName( 'si_text_idx', 'raw' ) ) . ")" );
 255                 $dbw->query( "CALL ctx_ddl.sync_index(" .
 256                         $dbw->addQuotes( $dbw->getDBname() . '.' . $dbw->tableName( 'si_title_idx', 'raw' ) ) . ")" );
 257         }
 258
 259         /**
 260          * Update a search index record's title only.
 261          * Title should be pre-processed.
 262          *
 263          * @param int $id
 264          * @param string $title
 265          */
 266         function updateTitle( $id, $title ) {
 267                 $dbw = wfGetDB( DB_MASTER );
 268
 269                 $dbw->update( 'searchindex',
 270                         array( 'si_title' => $title ),
 271                         array( 'si_page' => $id ),
 272                         'SearchOracle::updateTitle',
 273                         array() );
 274         }
 275
 276         public static function legalSearchChars() {
 277                 return "\"" . parent::legalSearchChars();
 278         }
 279 }