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 /**
  28  * Query module to perform full text search within wiki titles and content
  29  *
  30  * @ingroup API
  31  */
  32 class ApiQuerySearch extends ApiQueryGeneratorBase {
  33
  34         /**
  35          * When $wgSearchType is null, $wgSearchAlternatives[0] is null. Null isn't
  36          * a valid option for an array for PARAM_TYPE, so we'll use a fake name
  37          * that can't possibly be a class name and describes what the null behavior
  38          * does
  39          */
  40         const BACKEND_NULL_PARAM = 'database-backed';
  41
  42         public function __construct( $query, $moduleName ) {
  43                 parent::__construct( $query, $moduleName, 'sr' );
  44         }
  45
  46         public function execute() {
  47                 $this->run();
  48         }
  49
  50         public function executeGenerator( $resultPageSet ) {
  51                 $this->run( $resultPageSet );
  52         }
  53
  54         /**
  55          * @param $resultPageSet ApiPageSet
  56          * @return void
  57          */
  58         private function run( $resultPageSet = null ) {
  59                 global $wgContLang;
  60                 $params = $this->extractRequestParams();
  61
  62                 // Extract parameters
  63                 $limit = $params['limit'];
  64                 $query = $params['search'];
  65                 $what = $params['what'];
  66                 $searchInfo = array_flip( $params['info'] );
  67                 $prop = array_flip( $params['prop'] );
  68
  69                 // Create search engine instance and set options
  70                 $search = isset( $params['backend'] ) && $params['backend'] != self::BACKEND_NULL_PARAM ?
  71                         SearchEngine::create( $params['backend'] ) : SearchEngine::create();
  72                 $search->setLimitOffset( $limit + 1, $params['offset'] );
  73                 $search->setNamespaces( $params['namespace'] );
  74                 $search->showRedirects = $params['redirects'];
  75
  76                 $query = $search->transformSearchTerm( $query );
  77                 $query = $search->replacePrefixes( $query );
  78
  79                 // Perform the actual search
  80                 if ( $what == 'text' ) {
  81                         $matches = $search->searchText( $query );
  82                 } elseif ( $what == 'title' ) {
  83                         $matches = $search->searchTitle( $query );
  84                 } elseif ( $what == 'nearmatch' ) {
  85                         $matches = SearchEngine::getNearMatchResultSet( $query );
  86                 } else {
  87                         // We default to title searches; this is a terrible legacy
  88                         // of the way we initially set up the MySQL fulltext-based
  89                         // search engine with separate title and text fields.
  90                         // In the future, the default should be for a combined index.
  91                         $what = 'title';
  92                         $matches = $search->searchTitle( $query );
  93
  94                         // Not all search engines support a separate title search,
  95                         // for instance the Lucene-based engine we use on Wikipedia.
  96                         // In this case, fall back to full-text search (which will
  97                         // include titles in it!)
  98                         if ( is_null( $matches ) ) {
  99                                 $what = 'text';
 100                                 $matches = $search->searchText( $query );
 101                         }
 102                 }
 103                 if ( is_null( $matches ) ) {
 104                         $this->dieUsage( "{$what} search is disabled", "search-{$what}-disabled" );
 105                 } elseif ( $matches instanceof Status && !$matches->isGood() ) {
 106                         $this->dieUsage( $matches->getWikiText(), 'search-error' );
 107                 }
 108
 109                 $apiResult = $this->getResult();
 110                 // Add search meta data to result
 111                 if ( isset( $searchInfo['totalhits'] ) ) {
 112                         $totalhits = $matches->getTotalHits();
 113                         if ( $totalhits !== null ) {
 114                                 $apiResult->addValue( array( 'query', 'searchinfo' ),
 115                                                 'totalhits', $totalhits );
 116                         }
 117                 }
 118                 if ( isset( $searchInfo['suggestion'] ) && $matches->hasSuggestion() ) {
 119                         $apiResult->addValue( array( 'query', 'searchinfo' ),
 120                                                 'suggestion', $matches->getSuggestionQuery() );
 121                 }
 122
 123                 // Add the search results to the result
 124                 $terms = $wgContLang->convertForSearchResult( $matches->termMatches() );
 125                 $titles = array();
 126                 $count = 0;
 127                 $result = $matches->next();
 128
 129                 while ( $result ) {
 130                         if ( ++ $count > $limit ) {
 131                                 // We've reached the one extra which shows that there are additional items to be had. Stop here...
 132                                 $this->setContinueEnumParameter( 'offset', $params['offset'] + $params['limit'] );
 133                                 break;
 134                         }
 135
 136                         // Silently skip broken and missing titles
 137                         if ( $result->isBrokenTitle() || $result->isMissingRevision() ) {
 138                                 $result = $matches->next();
 139                                 continue;
 140                         }
 141
 142                         $title = $result->getTitle();
 143                         if ( is_null( $resultPageSet ) ) {
 144                                 $vals = array();
 145                                 ApiQueryBase::addTitleInfo( $vals, $title );
 146
 147                                 if ( isset( $prop['snippet'] ) ) {
 148                                         $vals['snippet'] = $result->getTextSnippet( $terms );
 149                                 }
 150                                 if ( isset( $prop['size'] ) ) {
 151                                         $vals['size'] = $result->getByteSize();
 152                                 }
 153                                 if ( isset( $prop['wordcount'] ) ) {
 154                                         $vals['wordcount'] = $result->getWordCount();
 155                                 }
 156                                 if ( isset( $prop['timestamp'] ) ) {
 157                                         $vals['timestamp'] = wfTimestamp( TS_ISO_8601, $result->getTimestamp() );
 158                                 }
 159                                 if ( !is_null( $result->getScore() ) && isset( $prop['score'] ) ) {
 160                                         $vals['score'] = $result->getScore();
 161                                 }
 162                                 if ( isset( $prop['titlesnippet'] ) ) {
 163                                         $vals['titlesnippet'] = $result->getTitleSnippet( $terms );
 164                                 }
 165                                 if ( !is_null( $result->getRedirectTitle() ) ) {
 166                                         if ( isset( $prop['redirecttitle'] ) ) {
 167                                                 $vals['redirecttitle'] = $result->getRedirectTitle();
 168                                         }
 169                                         if ( isset( $prop['redirectsnippet'] ) ) {
 170                                                 $vals['redirectsnippet'] = $result->getRedirectSnippet( $terms );
 171                                         }
 172                                 }
 173                                 if ( !is_null( $result->getSectionTitle() ) ) {
 174                                         if ( isset( $prop['sectiontitle'] ) ) {
 175                                                 $vals['sectiontitle'] = $result->getSectionTitle()->getFragment();
 176                                         }
 177                                         if ( isset( $prop['sectionsnippet'] ) ) {
 178                                                 $vals['sectionsnippet'] = $result->getSectionSnippet();
 179                                         }
 180                                 }
 181                                 if ( isset( $prop['hasrelated'] ) && $result->hasRelated() ) {
 182                                         $vals['hasrelated'] = '';
 183                                 }
 184
 185                                 // Add item to results and see whether it fits
 186                                 $fit = $apiResult->addValue( array( 'query', $this->getModuleName() ),
 187                                                 null, $vals );
 188                                 if ( !$fit ) {
 189                                         $this->setContinueEnumParameter( 'offset', $params['offset'] + $count - 1 );
 190                                         break;
 191                                 }
 192                         } else {
 193                                 $titles[] = $title;
 194                         }
 195
 196                         $result = $matches->next();
 197                 }
 198
 199                 if ( is_null( $resultPageSet ) ) {
 200                         $apiResult->setIndexedTagName_internal( array(
 201                                                 'query', $this->getModuleName()
 202                                         ), 'p' );
 203                 } else {
 204                         $resultPageSet->populateFromTitles( $titles );
 205                 }
 206         }
 207
 208         public function getCacheMode( $params ) {
 209                 return 'public';
 210         }
 211
 212         public function getAllowedParams() {
 213                 global $wgSearchType;
 214
 215                 $params = array(
 216                         'search' => array(
 217                                 ApiBase::PARAM_TYPE => 'string',
 218                                 ApiBase::PARAM_REQUIRED => true
 219                         ),
 220                         'namespace' => array(
 221                                 ApiBase::PARAM_DFLT => NS_MAIN,
 222                                 ApiBase::PARAM_TYPE => 'namespace',
 223                                 ApiBase::PARAM_ISMULTI => true,
 224                         ),
 225                         'what' => array(
 226                                 ApiBase::PARAM_DFLT => null,
 227                                 ApiBase::PARAM_TYPE => array(
 228                                         'title',
 229                                         'text',
 230                                         'nearmatch',
 231                                 )
 232                         ),
 233                         'info' => array(
 234                                 ApiBase::PARAM_DFLT => 'totalhits|suggestion',
 235                                 ApiBase::PARAM_TYPE => array(
 236                                         'totalhits',
 237                                         'suggestion',
 238                                 ),
 239                                 ApiBase::PARAM_ISMULTI => true,
 240                         ),
 241                         'prop' => array(
 242                                 ApiBase::PARAM_DFLT => 'size|wordcount|timestamp|snippet',
 243                                 ApiBase::PARAM_TYPE => array(
 244                                         'size',
 245                                         'wordcount',
 246                                         'timestamp',
 247                                         'score',
 248                                         'snippet',
 249                                         'titlesnippet',
 250                                         'redirecttitle',
 251                                         'redirectsnippet',
 252                                         'sectiontitle',
 253                                         'sectionsnippet',
 254                                         'hasrelated',
 255                                 ),
 256                                 ApiBase::PARAM_ISMULTI => true,
 257                         ),
 258                         'redirects' => false,
 259                         'offset' => 0,
 260                         'limit' => array(
 261                                 ApiBase::PARAM_DFLT => 10,
 262                                 ApiBase::PARAM_TYPE => 'limit',
 263                                 ApiBase::PARAM_MIN => 1,
 264                                 ApiBase::PARAM_MAX => ApiBase::LIMIT_SML1,
 265                                 ApiBase::PARAM_MAX2 => ApiBase::LIMIT_SML2
 266                         )
 267                 );
 268
 269                 $alternatives = SearchEngine::getSearchTypes();
 270                 if ( count( $alternatives ) > 1 ) {
 271                         if ( $alternatives[0] === null ) {
 272                                 $alternatives[0] = self::BACKEND_NULL_PARAM;
 273                         }
 274                         $params['backend'] = array(
 275                                 ApiBase::PARAM_DFLT => $wgSearchType,
 276                                 ApiBase::PARAM_TYPE => $alternatives,
 277                         );
 278                 }
 279
 280                 return $params;
 281         }
 282
 283         public function getParamDescription() {
 284                 $descriptions = array(
 285                         'search' => 'Search for all page titles (or content) that has this value',
 286                         'namespace' => 'The namespace(s) to enumerate',
 287                         'what' => 'Search inside the text or titles',
 288                         'info' => 'What metadata to return',
 289                         'prop' => array(
 290                                 'What properties to return',
 291                                 ' size             - Adds the size of the page in bytes',
 292                                 ' wordcount        - Adds the word count of the page',
 293                                 ' timestamp        - Adds the timestamp of when the page was last edited',
 294                                 ' score            - Adds the score (if any) from the search engine',
 295                                 ' snippet          - Adds a parsed snippet of the page',
 296                                 ' titlesnippet     - Adds a parsed snippet of the page title',
 297                                 ' redirectsnippet  - Adds a parsed snippet of the redirect title',
 298                                 ' redirecttitle    - Adds the title of the matching redirect',
 299                                 ' sectionsnippet   - Adds a parsed snippet of the matching section title',
 300                                 ' sectiontitle     - Adds the title of the matching section',
 301                                 ' hasrelated       - Indicates whether a related search is available',
 302                         ),
 303                         'redirects' => 'Include redirect pages in the search',
 304                         'offset' => 'Use this value to continue paging (return by query)',
 305                         'limit' => 'How many total pages to return'
 306                 );
 307
 308                 if ( count( SearchEngine::getSearchTypes() ) > 1 ) {
 309                         $descriptions['backend'] = 'Which search backend to use, if not the default';
 310                 }
 311
 312                 return $descriptions;
 313         }
 314
 315         public function getResultProperties() {
 316                 return array(
 317                         '' => array(
 318                                 'ns' => 'namespace',
 319                                 'title' => 'string'
 320                         ),
 321                         'snippet' => array(
 322                                 'snippet' => 'string'
 323                         ),
 324                         'size' => array(
 325                                 'size' => 'integer'
 326                         ),
 327                         'wordcount' => array(
 328                                 'wordcount' => 'integer'
 329                         ),
 330                         'timestamp' => array(
 331                                 'timestamp' => 'timestamp'
 332                         ),
 333                         'score' => array(
 334                                 'score' => array(
 335                                         ApiBase::PROP_TYPE => 'string',
 336                                         ApiBase::PROP_NULLABLE => true
 337                                 )
 338                         ),
 339                         'titlesnippet' => array(
 340                                 'titlesnippet' => 'string'
 341                         ),
 342                         'redirecttitle' => array(
 343                                 'redirecttitle' => array(
 344                                         ApiBase::PROP_TYPE => 'string',
 345                                         ApiBase::PROP_NULLABLE => true
 346                                 )
 347                         ),
 348                         'redirectsnippet' => array(
 349                                 'redirectsnippet' => array(
 350                                         ApiBase::PROP_TYPE => 'string',
 351                                         ApiBase::PROP_NULLABLE => true
 352                                 )
 353                         ),
 354                         'sectiontitle' => array(
 355                                 'sectiontitle' => array(
 356                                         ApiBase::PROP_TYPE => 'string',
 357                                         ApiBase::PROP_NULLABLE => true
 358                                 )
 359                         ),
 360                         'sectionsnippet' => array(
 361                                 'sectionsnippet' => array(
 362                                         ApiBase::PROP_TYPE => 'string',
 363                                         ApiBase::PROP_NULLABLE => true
 364                                 )
 365                         ),
 366                         'hasrelated' => array(
 367                                 'hasrelated' => 'boolean'
 368                         )
 369                 );
 370         }
 371
 372         public function getDescription() {
 373                 return 'Perform a full text search';
 374         }
 375
 376         public function getPossibleErrors() {
 377                 return array_merge( parent::getPossibleErrors(), array(
 378                         array( 'code' => 'search-text-disabled', 'info' => 'text search is disabled' ),
 379                         array( 'code' => 'search-title-disabled', 'info' => 'title search is disabled' ),
 380                         array( 'code' => 'search-error', 'info' => 'search error has occurred' ),
 381                 ) );
 382         }
 383
 384         public function getExamples() {
 385                 return array(
 386                         'api.php?action=query&list=search&srsearch=meaning',
 387                         'api.php?action=query&list=search&srwhat=text&srsearch=meaning',
 388                         'api.php?action=query&generator=search&gsrsearch=meaning&prop=info',
 389                 );
 390         }
 391
 392         public function getHelpUrls() {
 393                 return 'https://www.mediawiki.org/wiki/API:Search';
 394         }
 395 }