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         public function __construct( $query, $moduleName ) {
  35                 parent::__construct( $query, $moduleName, 'sr' );
  36         }
  37
  38         public function execute() {
  39                 $this->run();
  40         }
  41
  42         public function executeGenerator( $resultPageSet ) {
  43                 $this->run( $resultPageSet );
  44         }
  45
  46         /**
  47          * @param $resultPageSet ApiPageSet
  48          * @return void
  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                 // Create search engine instance and set options
  62                 $search = SearchEngine::create();
  63                 $search->setLimitOffset( $limit + 1, $params['offset'] );
  64                 $search->setNamespaces( $params['namespace'] );
  65                 $search->showRedirects = $params['redirects'];
  66
  67                 $query = $search->transformSearchTerm( $query );
  68                 $query = $search->replacePrefixes( $query );
  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                 } elseif ( $what == 'nearmatch' ) {
  76                         $matches = SearchEngine::getNearMatchResultSet( $query );
  77                 } else {
  78                         // We default to title searches; this is a terrible legacy
  79                         // of the way we initially set up the MySQL fulltext-based
  80                         // search engine with separate title and text fields.
  81                         // In the future, the default should be for a combined index.
  82                         $what = 'title';
  83                         $matches = $search->searchTitle( $query );
  84
  85                         // Not all search engines support a separate title search,
  86                         // for instance the Lucene-based engine we use on Wikipedia.
  87                         // In this case, fall back to full-text search (which will
  88                         // include titles in it!)
  89                         if ( is_null( $matches ) ) {
  90                                 $what = 'text';
  91                                 $matches = $search->searchText( $query );
  92                         }
  93                 }
  94                 if ( is_null( $matches ) ) {
  95                         $this->dieUsage( "{$what} search is disabled", "search-{$what}-disabled" );
  96                 } elseif ( $matches instanceof Status && !$matches->isGood() ) {
  97                         $this->dieUsage( $matches->getWikiText(), 'search-error' );
  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                                 $result = $matches->next();
 130                                 continue;
 131                         }
 132
 133                         $title = $result->getTitle();
 134                         if ( is_null( $resultPageSet ) ) {
 135                                 $vals = array();
 136                                 ApiQueryBase::addTitleInfo( $vals, $title );
 137
 138                                 if ( isset( $prop['snippet'] ) ) {
 139                                         $vals['snippet'] = $result->getTextSnippet( $terms );
 140                                 }
 141                                 if ( isset( $prop['size'] ) ) {
 142                                         $vals['size'] = $result->getByteSize();
 143                                 }
 144                                 if ( isset( $prop['wordcount'] ) ) {
 145                                         $vals['wordcount'] = $result->getWordCount();
 146                                 }
 147                                 if ( isset( $prop['timestamp'] ) ) {
 148                                         $vals['timestamp'] = wfTimestamp( TS_ISO_8601, $result->getTimestamp() );
 149                                 }
 150                                 if ( !is_null( $result->getScore() ) && isset( $prop['score'] ) ) {
 151                                         $vals['score'] = $result->getScore();
 152                                 }
 153                                 if ( isset( $prop['titlesnippet'] ) ) {
 154                                         $vals['titlesnippet'] = $result->getTitleSnippet( $terms );
 155                                 }
 156                                 if ( !is_null( $result->getRedirectTitle() ) ) {
 157                                         if ( isset( $prop['redirecttitle'] ) ) {
 158                                                 $vals['redirecttitle'] = $result->getRedirectTitle();
 159                                         }
 160                                         if ( isset( $prop['redirectsnippet'] ) ) {
 161                                                 $vals['redirectsnippet'] = $result->getRedirectSnippet( $terms );
 162                                         }
 163                                 }
 164                                 if ( !is_null( $result->getSectionTitle() ) ) {
 165                                         if ( isset( $prop['sectiontitle'] ) ) {
 166                                                 $vals['sectiontitle'] = $result->getSectionTitle()->getFragment();
 167                                         }
 168                                         if ( isset( $prop['sectionsnippet'] ) ) {
 169                                                 $vals['sectionsnippet'] = $result->getSectionSnippet();
 170                                         }
 171                                 }
 172                                 if ( isset( $prop['hasrelated'] ) && $result->hasRelated() ) {
 173                                         $vals['hasrelated'] = '';
 174                                 }
 175
 176                                 // Add item to results and see whether it fits
 177                                 $fit = $apiResult->addValue( array( 'query', $this->getModuleName() ),
 178                                                 null, $vals );
 179                                 if ( !$fit ) {
 180                                         $this->setContinueEnumParameter( 'offset', $params['offset'] + $count - 1 );
 181                                         break;
 182                                 }
 183                         } else {
 184                                 $titles[] = $title;
 185                         }
 186
 187                         $result = $matches->next();
 188                 }
 189
 190                 if ( is_null( $resultPageSet ) ) {
 191                         $apiResult->setIndexedTagName_internal( array(
 192                                                 'query', $this->getModuleName()
 193                                         ), 'p' );
 194                 } else {
 195                         $resultPageSet->populateFromTitles( $titles );
 196                 }
 197         }
 198
 199         public function getCacheMode( $params ) {
 200                 return 'public';
 201         }
 202
 203         public function getAllowedParams() {
 204                 return array(
 205                         'search' => array(
 206                                 ApiBase::PARAM_TYPE => 'string',
 207                                 ApiBase::PARAM_REQUIRED => true
 208                         ),
 209                         'namespace' => array(
 210                                 ApiBase::PARAM_DFLT => NS_MAIN,
 211                                 ApiBase::PARAM_TYPE => 'namespace',
 212                                 ApiBase::PARAM_ISMULTI => true,
 213                         ),
 214                         'what' => array(
 215                                 ApiBase::PARAM_DFLT => null,
 216                                 ApiBase::PARAM_TYPE => array(
 217                                         'title',
 218                                         'text',
 219                                         'nearmatch',
 220                                 )
 221                         ),
 222                         'info' => array(
 223                                 ApiBase::PARAM_DFLT => 'totalhits|suggestion',
 224                                 ApiBase::PARAM_TYPE => array(
 225                                         'totalhits',
 226                                         'suggestion',
 227                                 ),
 228                                 ApiBase::PARAM_ISMULTI => true,
 229                         ),
 230                         'prop' => array(
 231                                 ApiBase::PARAM_DFLT => 'size|wordcount|timestamp|snippet',
 232                                 ApiBase::PARAM_TYPE => array(
 233                                         'size',
 234                                         'wordcount',
 235                                         'timestamp',
 236                                         'score',
 237                                         'snippet',
 238                                         'titlesnippet',
 239                                         'redirecttitle',
 240                                         'redirectsnippet',
 241                                         'sectiontitle',
 242                                         'sectionsnippet',
 243                                         'hasrelated',
 244                                 ),
 245                                 ApiBase::PARAM_ISMULTI => true,
 246                         ),
 247                         'redirects' => false,
 248                         'offset' => 0,
 249                         'limit' => array(
 250                                 ApiBase::PARAM_DFLT => 10,
 251                                 ApiBase::PARAM_TYPE => 'limit',
 252                                 ApiBase::PARAM_MIN => 1,
 253                                 ApiBase::PARAM_MAX => ApiBase::LIMIT_SML1,
 254                                 ApiBase::PARAM_MAX2 => ApiBase::LIMIT_SML2
 255                         )
 256                 );
 257         }
 258
 259         public function getParamDescription() {
 260                 return array(
 261                         'search' => 'Search for all page titles (or content) that has this value',
 262                         'namespace' => 'The namespace(s) to enumerate',
 263                         'what' => 'Search inside the text or titles',
 264                         'info' => 'What metadata to return',
 265                         'prop' => array(
 266                                 'What properties to return',
 267                                 ' size             - Adds the size of the page in bytes',
 268                                 ' wordcount        - Adds the word count of the page',
 269                                 ' timestamp        - Adds the timestamp of when the page was last edited',
 270                                 ' score            - Adds the score (if any) from the search engine',
 271                                 ' snippet          - Adds a parsed snippet of the page',
 272                                 ' titlesnippet     - Adds a parsed snippet of the page title',
 273                                 ' redirectsnippet  - Adds a parsed snippet of the redirect title',
 274                                 ' redirecttitle    - Adds the title of the matching redirect',
 275                                 ' sectionsnippet   - Adds a parsed snippet of the matching section title',
 276                                 ' sectiontitle     - Adds the title of the matching section',
 277                                 ' hasrelated       - Indicates whether a related search is available',
 278                         ),
 279                         'redirects' => 'Include redirect pages in the search',
 280                         'offset' => 'Use this value to continue paging (return by query)',
 281                         'limit' => 'How many total pages to return'
 282                 );
 283         }
 284
 285         public function getResultProperties() {
 286                 return array(
 287                         '' => array(
 288                                 'ns' => 'namespace',
 289                                 'title' => 'string'
 290                         ),
 291                         'snippet' => array(
 292                                 'snippet' => 'string'
 293                         ),
 294                         'size' => array(
 295                                 'size' => 'integer'
 296                         ),
 297                         'wordcount' => array(
 298                                 'wordcount' => 'integer'
 299                         ),
 300                         'timestamp' => array(
 301                                 'timestamp' => 'timestamp'
 302                         ),
 303                         'score' => array(
 304                                 'score' => array(
 305                                         ApiBase::PROP_TYPE => 'string',
 306                                         ApiBase::PROP_NULLABLE => true
 307                                 )
 308                         ),
 309                         'titlesnippet' => array(
 310                                 'titlesnippet' => 'string'
 311                         ),
 312                         'redirecttitle' => array(
 313                                 'redirecttitle' => array(
 314                                         ApiBase::PROP_TYPE => 'string',
 315                                         ApiBase::PROP_NULLABLE => true
 316                                 )
 317                         ),
 318                         'redirectsnippet' => array(
 319                                 'redirectsnippet' => array(
 320                                         ApiBase::PROP_TYPE => 'string',
 321                                         ApiBase::PROP_NULLABLE => true
 322                                 )
 323                         ),
 324                         'sectiontitle' => array(
 325                                 'sectiontitle' => array(
 326                                         ApiBase::PROP_TYPE => 'string',
 327                                         ApiBase::PROP_NULLABLE => true
 328                                 )
 329                         ),
 330                         'sectionsnippet' => array(
 331                                 'sectionsnippet' => array(
 332                                         ApiBase::PROP_TYPE => 'string',
 333                                         ApiBase::PROP_NULLABLE => true
 334                                 )
 335                         ),
 336                         'hasrelated' => array(
 337                                 'hasrelated' => 'boolean'
 338                         )
 339                 );
 340         }
 341
 342         public function getDescription() {
 343                 return 'Perform a full text search';
 344         }
 345
 346         public function getPossibleErrors() {
 347                 return array_merge( parent::getPossibleErrors(), array(
 348                         array( 'code' => 'search-text-disabled', 'info' => 'text search is disabled' ),
 349                         array( 'code' => 'search-title-disabled', 'info' => 'title search is disabled' ),
 350                         array( 'code' => 'search-error', 'info' => 'search error has occurred' ),
 351                 ) );
 352         }
 353
 354         public function getExamples() {
 355                 return array(
 356                         'api.php?action=query&list=search&srsearch=meaning',
 357                         'api.php?action=query&list=search&srwhat=text&srsearch=meaning',
 358                         'api.php?action=query&generator=search&gsrsearch=meaning&prop=info',
 359                 );
 360         }
 361
 362         public function getHelpUrls() {
 363                 return 'https://www.mediawiki.org/wiki/API:Search';
 364         }
 365 }