includes/api/ApiQuerySearch.php

   1 <?php
   2
   3 /**
   4  * Created on July 30, 2007
   5  *
   6  * API for MediaWiki 1.8+
   7  *
   8  * Copyright © 2007 Yuri Astrakhan <Firstname><Lastname>@gmail.com
   9  *
  10  * This program is free software; you can redistribute it and/or modify
  11  * it under the terms of the GNU General Public License as published by
  12  * the Free Software Foundation; either version 2 of the License, or
  13  * (at your option) any later version.
  14  *
  15  * This program is distributed in the hope that it will be useful,
  16  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  17  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  18  * GNU General Public License for more details.
  19  *
  20  * You should have received a copy of the GNU General Public License along
  21  * with this program; if not, write to the Free Software Foundation, Inc.,
  22  * 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
  23  * http://www.gnu.org/copyleft/gpl.html
  24  */
  25
  26 if ( !defined( 'MEDIAWIKI' ) ) {
  27         // Eclipse helper - will be ignored in production
  28         require_once( 'ApiQueryBase.php' );
  29 }
  30
  31 /**
  32  * Query module to perform full text search within wiki titles and content
  33  *
  34  * @ingroup API
  35  */
  36 class ApiQuerySearch extends ApiQueryGeneratorBase {
  37
  38         public function __construct( $query, $moduleName ) {
  39                 parent::__construct( $query, $moduleName, 'sr' );
  40         }
  41
  42         public function execute() {
  43                 $this->run();
  44         }
  45
  46         public function executeGenerator( $resultPageSet ) {
  47                 $this->run( $resultPageSet );
  48         }
  49
  50         private function run( $resultPageSet = null ) {
  51                 global $wgContLang;
  52                 $params = $this->extractRequestParams();
  53
  54                 // Extract parameters
  55                 $limit = $params['limit'];
  56                 $query = $params['search'];
  57                 $what = $params['what'];
  58                 $searchInfo = array_flip( $params['info'] );
  59                 $prop = array_flip( $params['prop'] );
  60
  61                 if ( strval( $query ) === '' ) {
  62                         $this->dieUsage( 'empty search string is not allowed', 'param-search' );
  63                 }
  64
  65                 // Create search engine instance and set options
  66                 $search = SearchEngine::create();
  67                 $search->setLimitOffset( $limit + 1, $params['offset'] );
  68                 $search->setNamespaces( $params['namespace'] );
  69                 $search->showRedirects = $params['redirects'];
  70
  71                 // Perform the actual search
  72                 if ( $what == 'text' ) {
  73                         $matches = $search->searchText( $query );
  74                 } elseif ( $what == 'title' ) {
  75                         $matches = $search->searchTitle( $query );
  76                 } else {
  77                         // We default to title searches; this is a terrible legacy
  78                         // of the way we initially set up the MySQL fulltext-based
  79                         // search engine with separate title and text fields.
  80                         // In the future, the default should be for a combined index.
  81                         $what = 'title';
  82                         $matches = $search->searchTitle( $query );
  83
  84                         // Not all search engines support a separate title search,
  85                         // for instance the Lucene-based engine we use on Wikipedia.
  86                         // In this case, fall back to full-text search (which will
  87                         // include titles in it!)
  88                         if ( is_null( $matches ) ) {
  89                                 $what = 'text';
  90                                 $matches = $search->searchText( $query );
  91                         }
  92                 }
  93                 if ( is_null( $matches ) ) {
  94                         $this->dieUsage( "{$what} search is disabled", "search-{$what}-disabled" );
  95                 }
  96
  97                 // Add search meta data to result
  98                 if ( isset( $searchInfo['totalhits'] ) ) {
  99                         $totalhits = $matches->getTotalHits();
 100                         if ( $totalhits !== null ) {
 101                                 $this->getResult()->addValue( array( 'query', 'searchinfo' ),
 102                                                 'totalhits', $totalhits );
 103                         }
 104                 }
 105                 if ( isset( $searchInfo['suggestion'] ) && $matches->hasSuggestion() ) {
 106                         $this->getResult()->addValue( array( 'query', 'searchinfo' ),
 107                                                 'suggestion', $matches->getSuggestionQuery() );
 108                 }
 109
 110                 // Add the search results to the result
 111                 $terms = $wgContLang->convertForSearchResult( $matches->termMatches() );
 112                 $titles = array();
 113                 $count = 0;
 114                 while ( $result = $matches->next() ) {
 115                         if ( ++ $count > $limit ) {
 116                                 // We've reached the one extra which shows that there are additional items to be had. Stop here...
 117                                 $this->setContinueEnumParameter( 'offset', $params['offset'] + $params['limit'] );
 118                                 break;
 119                         }
 120
 121                         // Silently skip broken and missing titles
 122                         if ( $result->isBrokenTitle() || $result->isMissingRevision() ) {
 123                                 continue;
 124                         }
 125
 126                         $title = $result->getTitle();
 127                         if ( is_null( $resultPageSet ) ) {
 128                                 $vals = array();
 129                                 ApiQueryBase::addTitleInfo( $vals, $title );
 130
 131                                 if ( isset( $prop['snippet'] ) ) {
 132                                         $vals['snippet'] = $result->getTextSnippet( $terms );
 133                                 }
 134                                 if ( isset( $prop['size'] ) ) {
 135                                         $vals['size'] = $result->getByteSize();
 136                                 }
 137                                 if ( isset( $prop['wordcount'] ) ) {
 138                                         $vals['wordcount'] = $result->getWordCount();
 139                                 }
 140                                 if ( isset( $prop['timestamp'] ) ) {
 141                                         $vals['timestamp'] = wfTimestamp( TS_ISO_8601, $result->getTimestamp() );
 142                                 }
 143
 144                                 // Add item to results and see whether it fits
 145                                 $fit = $this->getResult()->addValue( array( 'query', $this->getModuleName() ),
 146                                                 null, $vals );
 147                                 if ( !$fit ) {
 148                                         $this->setContinueEnumParameter( 'offset', $params['offset'] + $count - 1 );
 149                                         break;
 150                                 }
 151                         } else {
 152                                 $titles[] = $title;
 153                         }
 154                 }
 155
 156                 if ( is_null( $resultPageSet ) ) {
 157                         $this->getResult()->setIndexedTagName_internal( array(
 158                                                 'query', $this->getModuleName()
 159                                         ), 'p' );
 160                 } else {
 161                         $resultPageSet->populateFromTitles( $titles );
 162                 }
 163         }
 164
 165         public function getAllowedParams() {
 166                 return array(
 167                         'search' => null,
 168                         'namespace' => array(
 169                                 ApiBase::PARAM_DFLT => 0,
 170                                 ApiBase::PARAM_TYPE => 'namespace',
 171                                 ApiBase::PARAM_ISMULTI => true,
 172                         ),
 173                         'what' => array(
 174                                 ApiBase::PARAM_DFLT => null,
 175                                 ApiBase::PARAM_TYPE => array(
 176                                         'title',
 177                                         'text',
 178                                 )
 179                         ),
 180                         'info' => array(
 181                                 ApiBase::PARAM_DFLT => 'totalhits|suggestion',
 182                                 ApiBase::PARAM_TYPE => array(
 183                                         'totalhits',
 184                                         'suggestion',
 185                                 ),
 186                                 ApiBase::PARAM_ISMULTI => true,
 187                         ),
 188                         'prop' => array(
 189                                 ApiBase::PARAM_DFLT => 'size|wordcount|timestamp|snippet',
 190                                 ApiBase::PARAM_TYPE => array(
 191                                         'size',
 192                                         'wordcount',
 193                                         'timestamp',
 194                                         'snippet',
 195                                 ),
 196                                 ApiBase::PARAM_ISMULTI => true,
 197                         ),
 198                         'redirects' => false,
 199                         'offset' => 0,
 200                         'limit' => array(
 201                                 ApiBase::PARAM_DFLT => 10,
 202                                 ApiBase::PARAM_TYPE => 'limit',
 203                                 ApiBase::PARAM_MIN => 1,
 204                                 ApiBase::PARAM_MAX => ApiBase::LIMIT_SML1,
 205                                 ApiBase::PARAM_MAX2 => ApiBase::LIMIT_SML2
 206                         )
 207                 );
 208         }
 209
 210         public function getParamDescription() {
 211                 return array(
 212                         'search' => 'Search for all page titles (or content) that has this value.',
 213                         'namespace' => 'The namespace(s) to enumerate.',
 214                         'what' => 'Search inside the text or titles.',
 215                         'info' => 'What metadata to return.',
 216                         'prop' => 'What properties to return.',
 217                         'redirects' => 'Include redirect pages in the search.',
 218                         'offset' => 'Use this value to continue paging (return by query)',
 219                         'limit' => 'How many total pages to return.'
 220                 );
 221         }
 222
 223         public function getDescription() {
 224                 return 'Perform a full text search';
 225         }
 226
 227         public function getPossibleErrors() {
 228                 return array_merge( parent::getPossibleErrors(), array(
 229                         array( 'code' => 'param-search', 'info' => 'empty search string is not allowed' ),
 230                         array( 'code' => 'search-text-disabled', 'info' => 'text search is disabled' ),
 231                         array( 'code' => 'search-title-disabled', 'info' => 'title search is disabled' ),
 232                 ) );
 233         }
 234
 235         protected function getExamples() {
 236                 return array(
 237                         'api.php?action=query&list=search&srsearch=meaning',
 238                         'api.php?action=query&list=search&srwhat=text&srsearch=meaning',
 239                         'api.php?action=query&generator=search&gsrsearch=meaning&prop=info',
 240                 );
 241         }
 242
 243         public function getVersion() {
 244                 return __CLASS__ . ': $Id$';
 245         }
 246 }