includes/api/ApiQuerySearch.php

   1 <?php
   2
   3 /*
   4  * Created on July 30, 2007
   5  *
   6  * API for MediaWiki 1.8+
   7  *
   8  * Copyright (C) 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                 // Create search engine instance and set options
  65                 $search = SearchEngine::create();
  66                 $search->setLimitOffset( $limit+1, $params['offset'] );
  67                 $search->setNamespaces( $params['namespace'] );
  68                 $search->showRedirects = $params['redirects'];
  69
  70                 // Perform the actual search
  71                 if ( $what == 'text' ) {
  72                         $matches = $search->searchText( $query );
  73                 } elseif( $what == 'title' ) {
  74                         $matches = $search->searchTitle( $query );
  75                 } else {
  76                         // We default to title searches; this is a terrible legacy
  77                         // of the way we initially set up the MySQL fulltext-based
  78                         // search engine with separate title and text fields.
  79                         // In the future, the default should be for a combined index.
  80                         $what = 'title';
  81                         $matches = $search->searchTitle( $query );
  82
  83                         // Not all search engines support a separate title search,
  84                         // for instance the Lucene-based engine we use on Wikipedia.
  85                         // In this case, fall back to full-text search (which will
  86                         // include titles in it!)
  87                         if( is_null( $matches ) ) {
  88                                 $what = 'text';
  89                                 $matches = $search->searchText( $query );
  90                         }
  91                 }
  92                 if ( is_null( $matches ) )
  93                         $this->dieUsage( "{$what} search is disabled", "search-{$what}-disabled" );
  94
  95                 // Add search meta data to result
  96                 if ( isset( $searchInfo['totalhits'] ) ) {
  97                         $totalhits = $matches->getTotalHits();
  98                         if( $totalhits !== null ) {
  99                                 $this->getResult()->addValue( array( 'query', 'searchinfo' ),
 100                                                 'totalhits', $totalhits );
 101                         }
 102                 }
 103                 if ( isset( $searchInfo['suggestion'] ) && $matches->hasSuggestion() ) {
 104                         $this->getResult()->addValue( array( 'query', 'searchinfo' ),
 105                                                 'suggestion', $matches->getSuggestionQuery() );
 106                 }
 107
 108                 // Add the search results to the result
 109                 $terms = $wgContLang->convertForSearchResult( $matches->termMatches() );
 110                 $titles = array ();
 111                 $count = 0;
 112                 while ( $result = $matches->next() ) {
 113                         if ( ++ $count > $limit ) {
 114                                 // We've reached the one extra which shows that there are additional items to be had. Stop here...
 115                                 $this->setContinueEnumParameter( 'offset', $params['offset'] + $params['limit'] );
 116                                 break;
 117                         }
 118
 119                         // Silently skip broken and missing titles
 120                         if ( $result->isBrokenTitle() || $result->isMissingRevision() )
 121                                 continue;
 122
 123                         $title = $result->getTitle();
 124                         if ( is_null( $resultPageSet ) ) {
 125                                 $vals = array();
 126                                 ApiQueryBase::addTitleInfo( $vals, $title );
 127
 128                                 if ( isset( $prop['snippet'] ) )
 129                                         $vals['snippet'] = $result->getTextSnippet( $terms );
 130                                 if ( isset( $prop['size'] ) )
 131                                         $vals['size'] = $result->getByteSize();
 132                                 if ( isset( $prop['wordcount'] ) )
 133                                         $vals['wordcount'] = $result->getWordCount();
 134                                 if ( isset( $prop['timestamp'] ) )
 135                                         $vals['timestamp'] = wfTimestamp( TS_ISO_8601, $result->getTimestamp() );
 136
 137                                 // Add item to results and see whether it fits
 138                                 $fit = $this->getResult()->addValue( array( 'query', $this->getModuleName() ),
 139                                                 null, $vals );
 140                                 if ( !$fit ) {
 141                                         $this->setContinueEnumParameter( 'offset', $params['offset'] + $count - 1 );
 142                                         break;
 143                                 }
 144                         } else {
 145                                 $titles[] = $title;
 146                         }
 147                 }
 148
 149                 if ( is_null( $resultPageSet ) ) {
 150                         $this->getResult()->setIndexedTagName_internal( array(
 151                                                 'query', $this->getModuleName()
 152                                         ), 'p' );
 153                 } else {
 154                         $resultPageSet->populateFromTitles( $titles );
 155                 }
 156         }
 157
 158         public function getAllowedParams() {
 159                 return array (
 160                         'search' => null,
 161                         'namespace' => array (
 162                                 ApiBase :: PARAM_DFLT => 0,
 163                                 ApiBase :: PARAM_TYPE => 'namespace',
 164                                 ApiBase :: PARAM_ISMULTI => true,
 165                         ),
 166                         'what' => array (
 167                                 ApiBase :: PARAM_DFLT => null,
 168                                 ApiBase :: PARAM_TYPE => array (
 169                                         'title',
 170                                         'text',
 171                                 )
 172                         ),
 173                         'info' => array(
 174                                 ApiBase :: PARAM_DFLT => 'totalhits|suggestion',
 175                                 ApiBase :: PARAM_TYPE => array (
 176                                         'totalhits',
 177                                         'suggestion',
 178                                 ),
 179                                 ApiBase :: PARAM_ISMULTI => true,
 180                         ),
 181                         'prop' => array(
 182                                 ApiBase :: PARAM_DFLT => 'size|wordcount|timestamp|snippet',
 183                                 ApiBase :: PARAM_TYPE => array (
 184                                         'size',
 185                                         'wordcount',
 186                                         'timestamp',
 187                                         'snippet',
 188                                 ),
 189                                 ApiBase :: PARAM_ISMULTI => true,
 190                         ),
 191                         'redirects' => false,
 192                         'offset' => 0,
 193                         'limit' => array (
 194                                 ApiBase :: PARAM_DFLT => 10,
 195                                 ApiBase :: PARAM_TYPE => 'limit',
 196                                 ApiBase :: PARAM_MIN => 1,
 197                                 ApiBase :: PARAM_MAX => ApiBase :: LIMIT_BIG1,
 198                                 ApiBase :: PARAM_MAX2 => ApiBase :: LIMIT_BIG2
 199                         )
 200                 );
 201         }
 202
 203         public function getParamDescription() {
 204                 return array (
 205                         'search' => 'Search for all page titles (or content) that has this value.',
 206                         'namespace' => 'The namespace(s) to enumerate.',
 207                         'what' => 'Search inside the text or titles.',
 208                         'info' => 'What metadata to return.',
 209                         'prop' => 'What properties to return.',
 210                         'redirects' => 'Include redirect pages in the search.',
 211                         'offset' => 'Use this value to continue paging (return by query)',
 212                         'limit' => 'How many total pages to return.'
 213                 );
 214         }
 215
 216         public function getDescription() {
 217                 return 'Perform a full text search';
 218         }
 219
 220         protected function getExamples() {
 221                 return array (
 222                         'api.php?action=query&list=search&srsearch=meaning',
 223                         'api.php?action=query&list=search&srwhat=text&srsearch=meaning',
 224                         'api.php?action=query&generator=search&gsrsearch=meaning&prop=info',
 225                 );
 226         }
 227
 228         public function getVersion() {
 229                 return __CLASS__ . ': $Id$';
 230         }
 231 }