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  * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, 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                 } elseif ( $what == 'nearmatch' ) {
  77                         $query = str_replace( '_', ' ', $query );
  78                         $matches = SearchEngine::getNearMatchResultSet( $query );
  79                 } else {
  80                         // We default to title searches; this is a terrible legacy
  81                         // of the way we initially set up the MySQL fulltext-based
  82                         // search engine with separate title and text fields.
  83                         // In the future, the default should be for a combined index.
  84                         $what = 'title';
  85                         $matches = $search->searchTitle( $query );
  86
  87                         // Not all search engines support a separate title search,
  88                         // for instance the Lucene-based engine we use on Wikipedia.
  89                         // In this case, fall back to full-text search (which will
  90                         // include titles in it!)
  91                         if ( is_null( $matches ) ) {
  92                                 $what = 'text';
  93                                 $matches = $search->searchText( $query );
  94                         }
  95                 }
  96                 if ( is_null( $matches ) ) {
  97                         $this->dieUsage( "{$what} search is disabled", "search-{$what}-disabled" );
  98                 }
  99
 100                 // Add search meta data to result
 101                 if ( isset( $searchInfo['totalhits'] ) ) {
 102                         $totalhits = $matches->getTotalHits();
 103                         if ( $totalhits !== null ) {
 104                                 $this->getResult()->addValue( array( 'query', 'searchinfo' ),
 105                                                 'totalhits', $totalhits );
 106                         }
 107                 }
 108                 if ( isset( $searchInfo['suggestion'] ) && $matches->hasSuggestion() ) {
 109                         $this->getResult()->addValue( array( 'query', 'searchinfo' ),
 110                                                 'suggestion', $matches->getSuggestionQuery() );
 111                 }
 112
 113                 // Add the search results to the result
 114                 $terms = $wgContLang->convertForSearchResult( $matches->termMatches() );
 115                 $titles = array();
 116                 $count = 0;
 117                 while ( $result = $matches->next() ) {
 118                         if ( ++ $count > $limit ) {
 119                                 // We've reached the one extra which shows that there are additional items to be had. Stop here...
 120                                 $this->setContinueEnumParameter( 'offset', $params['offset'] + $params['limit'] );
 121                                 break;
 122                         }
 123
 124                         // Silently skip broken and missing titles
 125                         if ( $result->isBrokenTitle() || $result->isMissingRevision() ) {
 126                                 continue;
 127                         }
 128
 129                         $title = $result->getTitle();
 130                         if ( is_null( $resultPageSet ) ) {
 131                                 $vals = array();
 132                                 ApiQueryBase::addTitleInfo( $vals, $title );
 133
 134                                 if ( isset( $prop['snippet'] ) ) {
 135                                         $vals['snippet'] = $result->getTextSnippet( $terms );
 136                                 }
 137                                 if ( isset( $prop['size'] ) ) {
 138                                         $vals['size'] = $result->getByteSize();
 139                                 }
 140                                 if ( isset( $prop['wordcount'] ) ) {
 141                                         $vals['wordcount'] = $result->getWordCount();
 142                                 }
 143                                 if ( isset( $prop['timestamp'] ) ) {
 144                                         $vals['timestamp'] = wfTimestamp( TS_ISO_8601, $result->getTimestamp() );
 145                                 }
 146
 147                                 // Add item to results and see whether it fits
 148                                 $fit = $this->getResult()->addValue( array( 'query', $this->getModuleName() ),
 149                                                 null, $vals );
 150                                 if ( !$fit ) {
 151                                         $this->setContinueEnumParameter( 'offset', $params['offset'] + $count - 1 );
 152                                         break;
 153                                 }
 154                         } else {
 155                                 $titles[] = $title;
 156                         }
 157                 }
 158
 159                 if ( is_null( $resultPageSet ) ) {
 160                         $this->getResult()->setIndexedTagName_internal( array(
 161                                                 'query', $this->getModuleName()
 162                                         ), 'p' );
 163                 } else {
 164                         $resultPageSet->populateFromTitles( $titles );
 165                 }
 166         }
 167
 168         public function getAllowedParams() {
 169                 return array(
 170                         'search' => null,
 171                         'namespace' => array(
 172                                 ApiBase::PARAM_DFLT => 0,
 173                                 ApiBase::PARAM_TYPE => 'namespace',
 174                                 ApiBase::PARAM_ISMULTI => true,
 175                         ),
 176                         'what' => array(
 177                                 ApiBase::PARAM_DFLT => null,
 178                                 ApiBase::PARAM_TYPE => array(
 179                                         'title',
 180                                         'text',
 181                                         'nearmatch',
 182                                 )
 183                         ),
 184                         'info' => array(
 185                                 ApiBase::PARAM_DFLT => 'totalhits|suggestion',
 186                                 ApiBase::PARAM_TYPE => array(
 187                                         'totalhits',
 188                                         'suggestion',
 189                                 ),
 190                                 ApiBase::PARAM_ISMULTI => true,
 191                         ),
 192                         'prop' => array(
 193                                 ApiBase::PARAM_DFLT => 'size|wordcount|timestamp|snippet',
 194                                 ApiBase::PARAM_TYPE => array(
 195                                         'size',
 196                                         'wordcount',
 197                                         'timestamp',
 198                                         'snippet',
 199                                 ),
 200                                 ApiBase::PARAM_ISMULTI => true,
 201                         ),
 202                         'redirects' => false,
 203                         'offset' => 0,
 204                         'limit' => array(
 205                                 ApiBase::PARAM_DFLT => 10,
 206                                 ApiBase::PARAM_TYPE => 'limit',
 207                                 ApiBase::PARAM_MIN => 1,
 208                                 ApiBase::PARAM_MAX => ApiBase::LIMIT_SML1,
 209                                 ApiBase::PARAM_MAX2 => ApiBase::LIMIT_SML2
 210                         )
 211                 );
 212         }
 213
 214         public function getParamDescription() {
 215                 return array(
 216                         'search' => 'Search for all page titles (or content) that has this value',
 217                         'namespace' => 'The namespace(s) to enumerate',
 218                         'what' => 'Search inside the text or titles',
 219                         'info' => 'What metadata to return',
 220                         'prop' => array(
 221                                 'What properties to return',
 222                                 ' size    - Adds the size of the page in bytes',
 223                                 ' wordcount  - Adds the word count of the page',
 224                                 ' timestamp  - Adds the timestamp of when the page was last edited',
 225                                 ' snippet    - Adds a parsed snippet of the page',
 226                         ),
 227                         'redirects' => 'Include redirect pages in the search',
 228                         'offset' => 'Use this value to continue paging (return by query)',
 229                         'limit' => 'How many total pages to return'
 230                 );
 231         }
 232
 233         public function getDescription() {
 234                 return 'Perform a full text search';
 235         }
 236
 237         public function getPossibleErrors() {
 238                 return array_merge( parent::getPossibleErrors(), array(
 239                         array( 'code' => 'param-search', 'info' => 'empty search string is not allowed' ),
 240                         array( 'code' => 'search-text-disabled', 'info' => 'text search is disabled' ),
 241                         array( 'code' => 'search-title-disabled', 'info' => 'title search is disabled' ),
 242                 ) );
 243         }
 244
 245         protected function getExamples() {
 246                 return array(
 247                         'api.php?action=query&list=search&srsearch=meaning',
 248                         'api.php?action=query&list=search&srwhat=text&srsearch=meaning',
 249                         'api.php?action=query&generator=search&gsrsearch=meaning&prop=info',
 250                 );
 251         }
 252
 253         public function getVersion() {
 254                 return __CLASS__ . ': $Id$';
 255         }
 256 }