includes/api/ApiQuerySearch.php

   1 <?php
   2 /**
   3  *
   4  *
   5  * Created on July 30, 2007
   6  *
   7  * Copyright © 2007 Yuri Astrakhan <Firstname><Lastname>@gmail.com
   8  *
   9  * This program is free software; you can redistribute it and/or modify
  10  * it under the terms of the GNU General Public License as published by
  11  * the Free Software Foundation; either version 2 of the License, or
  12  * (at your option) any later version.
  13  *
  14  * This program is distributed in the hope that it will be useful,
  15  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  16  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  17  * GNU General Public License for more details.
  18  *
  19  * You should have received a copy of the GNU General Public License along
  20  * with this program; if not, write to the Free Software Foundation, Inc.,
  21  * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
  22  * http://www.gnu.org/copyleft/gpl.html
  23  *
  24  * @file
  25  */
  26
  27 if ( !defined( 'MEDIAWIKI' ) ) {
  28         // Eclipse helper - will be ignored in production
  29         require_once( 'ApiQueryBase.php' );
  30 }
  31
  32 /**
  33  * Query module to perform full text search within wiki titles and content
  34  *
  35  * @ingroup API
  36  */
  37 class ApiQuerySearch extends ApiQueryGeneratorBase {
  38
  39         public function __construct( $query, $moduleName ) {
  40                 parent::__construct( $query, $moduleName, 'sr' );
  41         }
  42
  43         public function execute() {
  44                 $this->run();
  45         }
  46
  47         public function executeGenerator( $resultPageSet ) {
  48                 $this->run( $resultPageSet );
  49         }
  50
  51         /**
  52          * @param $resultPageSet ApiPageSet
  53          * @return void
  54          */
  55         private function run( $resultPageSet = null ) {
  56                 global $wgContLang;
  57                 $params = $this->extractRequestParams();
  58
  59                 // Extract parameters
  60                 $limit = $params['limit'];
  61                 $query = $params['search'];
  62                 $what = $params['what'];
  63                 $searchInfo = array_flip( $params['info'] );
  64                 $prop = array_flip( $params['prop'] );
  65
  66                 // Create search engine instance and set options
  67                 $search = SearchEngine::create();
  68                 $search->setLimitOffset( $limit + 1, $params['offset'] );
  69                 $search->setNamespaces( $params['namespace'] );
  70                 $search->showRedirects = $params['redirects'];
  71
  72                 // Perform the actual search
  73                 if ( $what == 'text' ) {
  74                         $matches = $search->searchText( $query );
  75                 } elseif ( $what == 'title' ) {
  76                         $matches = $search->searchTitle( $query );
  77                 } elseif ( $what == 'nearmatch' ) {
  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                 $apiResult = $this->getResult();
 101                 // Add search meta data to result
 102                 if ( isset( $searchInfo['totalhits'] ) ) {
 103                         $totalhits = $matches->getTotalHits();
 104                         if ( $totalhits !== null ) {
 105                                 $apiResult->addValue( array( 'query', 'searchinfo' ),
 106                                                 'totalhits', $totalhits );
 107                         }
 108                 }
 109                 if ( isset( $searchInfo['suggestion'] ) && $matches->hasSuggestion() ) {
 110                         $apiResult->addValue( array( 'query', 'searchinfo' ),
 111                                                 'suggestion', $matches->getSuggestionQuery() );
 112                 }
 113
 114                 // Add the search results to the result
 115                 $terms = $wgContLang->convertForSearchResult( $matches->termMatches() );
 116                 $titles = array();
 117                 $count = 0;
 118                 $result = $matches->next();
 119
 120                 while ( $result ) {
 121                         if ( ++ $count > $limit ) {
 122                                 // We've reached the one extra which shows that there are additional items to be had. Stop here...
 123                                 $this->setContinueEnumParameter( 'offset', $params['offset'] + $params['limit'] );
 124                                 break;
 125                         }
 126
 127                         // Silently skip broken and missing titles
 128                         if ( $result->isBrokenTitle() || $result->isMissingRevision() ) {
 129                                 continue;
 130                         }
 131
 132                         $title = $result->getTitle();
 133                         if ( is_null( $resultPageSet ) ) {
 134                                 $vals = array();
 135                                 ApiQueryBase::addTitleInfo( $vals, $title );
 136
 137                                 if ( isset( $prop['snippet'] ) ) {
 138                                         $vals['snippet'] = $result->getTextSnippet( $terms );
 139                                 }
 140                                 if ( isset( $prop['size'] ) ) {
 141                                         $vals['size'] = $result->getByteSize();
 142                                 }
 143                                 if ( isset( $prop['wordcount'] ) ) {
 144                                         $vals['wordcount'] = $result->getWordCount();
 145                                 }
 146                                 if ( isset( $prop['timestamp'] ) ) {
 147                                         $vals['timestamp'] = wfTimestamp( TS_ISO_8601, $result->getTimestamp() );
 148                                 }
 149                                 if ( !is_null( $result->getScore() ) && isset( $prop['score'] ) ) {
 150                                         $vals['score'] = $result->getScore();
 151                                 }
 152                                 if ( isset( $prop['titlesnippet'] ) ) {
 153                                         $vals['titlesnippet'] = $result->getTitleSnippet( $terms );
 154                                 }
 155                                 if ( !is_null( $result->getRedirectTitle() ) ) {
 156                                         if ( isset( $prop['redirecttitle'] ) ) {
 157                                                 $vals['redirecttitle'] = $result->getRedirectTitle();
 158                                         }
 159                                         if ( isset( $prop['redirectsnippet'] ) ) {
 160                                                 $vals['redirectsnippet'] = $result->getRedirectSnippet( $terms );
 161                                         }
 162                                 }
 163                                 if ( !is_null( $result->getSectionTitle() ) ) {
 164                                         if ( isset( $prop['sectiontitle'] ) ) {
 165                                                 $vals['sectiontitle'] = $result->getSectionTitle()->getFragment();
 166                                         }
 167                                         if ( isset( $prop['sectionsnippet'] ) ) {
 168                                                 $vals['sectionsnippet'] = $result->getSectionSnippet();
 169                                         }
 170                                 }
 171                                 if ( isset( $prop['hasrelated'] ) && $result->hasRelated() ) {
 172                                         $vals['hasrelated'] = "";
 173                                 }
 174
 175                                 // Add item to results and see whether it fits
 176                                 $fit = $apiResult->addValue( array( 'query', $this->getModuleName() ),
 177                                                 null, $vals );
 178                                 if ( !$fit ) {
 179                                         $this->setContinueEnumParameter( 'offset', $params['offset'] + $count - 1 );
 180                                         break;
 181                                 }
 182                         } else {
 183                                 $titles[] = $title;
 184                         }
 185
 186                         $result = $matches->next();
 187                 }
 188
 189                 if ( is_null( $resultPageSet ) ) {
 190                         $apiResult->setIndexedTagName_internal( array(
 191                                                 'query', $this->getModuleName()
 192                                         ), 'p' );
 193                 } else {
 194                         $resultPageSet->populateFromTitles( $titles );
 195                 }
 196         }
 197
 198         public function getCacheMode( $params ) {
 199                 return 'public';
 200         }
 201
 202         public function getAllowedParams() {
 203                 return array(
 204                         'search' => array(
 205                                 ApiBase::PARAM_TYPE => 'string',
 206                                 ApiBase::PARAM_REQUIRED => true
 207                         ),
 208                         'namespace' => array(
 209                                 ApiBase::PARAM_DFLT => 0,
 210                                 ApiBase::PARAM_TYPE => 'namespace',
 211                                 ApiBase::PARAM_ISMULTI => true,
 212                         ),
 213                         'what' => array(
 214                                 ApiBase::PARAM_DFLT => null,
 215                                 ApiBase::PARAM_TYPE => array(
 216                                         'title',
 217                                         'text',
 218                                         'nearmatch',
 219                                 )
 220                         ),
 221                         'info' => array(
 222                                 ApiBase::PARAM_DFLT => 'totalhits|suggestion',
 223                                 ApiBase::PARAM_TYPE => array(
 224                                         'totalhits',
 225                                         'suggestion',
 226                                 ),
 227                                 ApiBase::PARAM_ISMULTI => true,
 228                         ),
 229                         'prop' => array(
 230                                 ApiBase::PARAM_DFLT => 'size|wordcount|timestamp|snippet',
 231                                 ApiBase::PARAM_TYPE => array(
 232                                         'size',
 233                                         'wordcount',
 234                                         'timestamp',
 235                                         'score',
 236                                         'snippet',
 237                                         'titlesnippet',
 238                                         'redirecttitle',
 239                                         'redirectsnippet',
 240                                         'sectiontitle',
 241                                         'sectionsnippet',
 242                                         'hasrelated',
 243                                 ),
 244                                 ApiBase::PARAM_ISMULTI => true,
 245                         ),
 246                         'redirects' => false,
 247                         'offset' => 0,
 248                         'limit' => array(
 249                                 ApiBase::PARAM_DFLT => 10,
 250                                 ApiBase::PARAM_TYPE => 'limit',
 251                                 ApiBase::PARAM_MIN => 1,
 252                                 ApiBase::PARAM_MAX => ApiBase::LIMIT_SML1,
 253                                 ApiBase::PARAM_MAX2 => ApiBase::LIMIT_SML2
 254                         )
 255                 );
 256         }
 257
 258         public function getParamDescription() {
 259                 return array(
 260                         'search' => 'Search for all page titles (or content) that has this value',
 261                         'namespace' => 'The namespace(s) to enumerate',
 262                         'what' => 'Search inside the text or titles',
 263                         'info' => 'What metadata to return',
 264                         'prop' => array(
 265                                 'What properties to return',
 266                                 ' size             - Adds the size of the page in bytes',
 267                                 ' wordcount        - Adds the word count of the page',
 268                                 ' timestamp        - Adds the timestamp of when the page was last edited',
 269                                 ' score            - Adds the score (if any) from the search engine',
 270                                 ' snippet          - Adds a parsed snippet of the page',
 271                                 ' titlesnippet     - Adds a parsed snippet of the page title',
 272                                 ' redirectsnippet  - Adds a parsed snippet of the redirect title',
 273                                 ' redirecttitle    - Adds the title of the matching redirect',
 274                                 ' sectionsnippet   - Adds a parsed snippet of the matching section title',
 275                                 ' sectiontitle     - Adds the title of the matching section',
 276                                 ' hasrelated       - Indicates whether a related search is available',
 277                         ),
 278                         'redirects' => 'Include redirect pages in the search',
 279                         'offset' => 'Use this value to continue paging (return by query)',
 280                         'limit' => 'How many total pages to return'
 281                 );
 282         }
 283
 284         public function getDescription() {
 285                 return 'Perform a full text search';
 286         }
 287
 288         public function getPossibleErrors() {
 289                 return array_merge( parent::getPossibleErrors(), array(
 290                         array( 'code' => 'search-text-disabled', 'info' => 'text search is disabled' ),
 291                         array( 'code' => 'search-title-disabled', 'info' => 'title search is disabled' ),
 292                 ) );
 293         }
 294
 295         protected function getExamples() {
 296                 return array(
 297                         'api.php?action=query&list=search&srsearch=meaning',
 298                         'api.php?action=query&list=search&srwhat=text&srsearch=meaning',
 299                         'api.php?action=query&generator=search&gsrsearch=meaning&prop=info',
 300                 );
 301         }
 302
 303         public function getVersion() {
 304                 return __CLASS__ . ': $Id$';
 305         }
 306 }