includes/api/ApiParse.php

   1 <?php
   2
   3 /**
   4  * Created on Dec 01, 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( "ApiBase.php" );
  29 }
  30
  31 /**
  32  * @ingroup API
  33  */
  34 class ApiParse extends ApiBase {
  35
  36         public function __construct( $main, $action ) {
  37                 parent::__construct( $main, $action );
  38         }
  39
  40         public function execute() {
  41                 // The data is hot but user-dependent, like page views, so we set vary cookies
  42                 $this->getMain()->setCacheMode( 'anon-public-user-private' );
  43
  44                 // Get parameters
  45                 $params = $this->extractRequestParams();
  46                 $text = $params['text'];
  47                 $title = $params['title'];
  48                 $page = $params['page'];
  49                 $pageid = $params['pageid'];
  50                 $oldid = $params['oldid'];
  51
  52                 if ( !is_null( $page ) && ( !is_null( $text ) || $title != 'API' ) ) {
  53                         $this->dieUsage( 'The page parameter cannot be used together with the text and title parameters', 'params' );
  54                 }
  55                 $prop = array_flip( $params['prop'] );
  56
  57                 if ( isset( $params['section'] ) ) {
  58                         $this->section = $params['section'];
  59                 } else {
  60                         $this->section = false;
  61                 }
  62
  63                 // The parser needs $wgTitle to be set, apparently the
  64                 // $title parameter in Parser::parse isn't enough *sigh*
  65                 global $wgParser, $wgUser, $wgTitle, $wgEnableParserCache, $wgLang;
  66
  67                 // Currently unnecessary, code to act as a safeguard against any change in current behaviour of uselang breaks
  68                 $oldLang = null;
  69                 if ( isset( $params['uselang'] ) && $params['uselang'] != $wgLang->getCode() ) {
  70                         $oldLang = $wgLang; // Backup wgLang
  71                         $wgLang = Language::factory( $params['uselang'] );
  72                 }
  73
  74                 $popts = new ParserOptions();
  75                 $popts->setTidy( true );
  76                 $popts->enableLimitReport();
  77                 $redirValues = null;
  78                 if ( !is_null( $oldid ) || !is_null( $pageid ) || !is_null( $page ) ) {
  79                         if ( !is_null( $oldid ) ) {
  80                                 // Don't use the parser cache
  81                                 $rev = Revision::newFromID( $oldid );
  82                                 if ( !$rev ) {
  83                                         $this->dieUsage( "There is no revision ID $oldid", 'missingrev' );
  84                                 }
  85                                 if ( !$rev->userCan( Revision::DELETED_TEXT ) ) {
  86                                         $this->dieUsage( "You don't have permission to view deleted revisions", 'permissiondenied' );
  87                                 }
  88
  89                                 $text = $rev->getText( Revision::FOR_THIS_USER );
  90                                 $titleObj = $rev->getTitle();
  91                                 $wgTitle = $titleObj;
  92
  93                                 if ( $this->section !== false ) {
  94                                         $text = $this->getSectionText( $text, 'r' . $rev );
  95                                 }
  96
  97                                 $p_result = $wgParser->parse( $text, $titleObj, $popts );
  98                         } else {
  99                                 if ( !is_null ( $pageid ) ) {
 100                                         $titleObj = Title::newFromID( $pageid );
 101
 102                                         if ( !$titleObj ) {
 103                                                 $this->dieUsageMsg( array( 'nosuchpageid', $pageid ) );
 104                                         }
 105                                 } else {
 106                                         if ( $params['redirects'] ) {
 107                                                 $req = new FauxRequest( array(
 108                                                         'action' => 'query',
 109                                                         'redirects' => '',
 110                                                         'titles' => $page
 111                                                 ) );
 112                                                 $main = new ApiMain( $req );
 113                                                 $main->execute();
 114                                                 $data = $main->getResultData();
 115                                                 $redirValues = @$data['query']['redirects'];
 116                                                 $to = $page;
 117                                                 foreach ( (array)$redirValues as $r ) {
 118                                                         $to = $r['to'];
 119                                                 }
 120                                         } else {
 121                                                 $to = $page;
 122                                         }
 123                                         $titleObj = Title::newFromText( $to );
 124                                         if ( !$titleObj ) {
 125                                                 $this->dieUsage( "The page you specified doesn't exist", 'missingtitle' );
 126                                         }
 127                                 }
 128                                 $wgTitle = $titleObj;
 129
 130                                 $articleObj = new Article( $titleObj );
 131                                 if ( isset( $prop['revid'] ) ) {
 132                                         $oldid = $articleObj->getRevIdFetched();
 133                                 }
 134
 135                                 if ( $this->section !== false ) {
 136                                         $text = $this->getSectionText( $text, !is_null ( $pageid ) ? 'page id ' . $pageid : $titleObj->getText() );
 137                                         $p_result = $wgParser->parse( $text, $titleObj, $popts );
 138                                 } else {
 139                                         // Try the parser cache first
 140                                         $p_result = false;
 141                                         $pcache = ParserCache::singleton();
 142                                         if ( $wgEnableParserCache ) {
 143                                                 $p_result = $pcache->get( $articleObj, $popts );
 144                                         }
 145                                         if ( !$p_result ) {
 146                                                 $p_result = $wgParser->parse( $articleObj->getContent(), $titleObj, $popts );
 147
 148                                                 if ( $wgEnableParserCache ) {
 149                                                         $pcache->save( $p_result, $articleObj, $popts );
 150                                                 }
 151                                         }
 152                                 }
 153                         }
 154                 } else {
 155                         $titleObj = Title::newFromText( $title );
 156                         if ( !$titleObj ) {
 157                                 $titleObj = Title::newFromText( 'API' );
 158                         }
 159                         $wgTitle = $titleObj;
 160
 161                         if ( $this->section !== false ) {
 162                                 $text = $this->getSectionText( $text, $titleObj->getText() );
 163                         }
 164
 165                         if ( $params['pst'] || $params['onlypst'] ) {
 166                                 $text = $wgParser->preSaveTransform( $text, $titleObj, $wgUser, $popts );
 167                         }
 168                         if ( $params['onlypst'] ) {
 169                                 // Build a result and bail out
 170                                 $result_array['text'] = array();
 171                                 $this->getResult()->setContent( $result_array['text'], $text );
 172                                 $this->getResult()->addValue( null, $this->getModuleName(), $result_array );
 173                                 return;
 174                         }
 175                         $p_result = $wgParser->parse( $text, $titleObj, $popts );
 176                 }
 177
 178                 // Return result
 179                 $result = $this->getResult();
 180                 $result_array = array();
 181                 if ( $params['redirects'] && !is_null( $redirValues ) ) {
 182                         $result_array['redirects'] = $redirValues;
 183                 }
 184
 185                 if ( isset( $prop['text'] ) ) {
 186                         $result_array['text'] = array();
 187                         $result->setContent( $result_array['text'], $p_result->getText() );
 188                 }
 189
 190                 if ( !is_null( $params['summary'] ) ) {
 191                         $result_array['parsedsummary'] = array();
 192                         $result->setContent( $result_array['parsedsummary'], $wgUser->getSkin()->formatComment( $params['summary'], $titleObj ) );
 193                 }
 194
 195                 if ( isset( $prop['langlinks'] ) ) {
 196                         $result_array['langlinks'] = $this->formatLangLinks( $p_result->getLanguageLinks() );
 197                 }
 198                 if ( isset( $prop['categories'] ) ) {
 199                         $result_array['categories'] = $this->formatCategoryLinks( $p_result->getCategories() );
 200                 }
 201                 if ( isset( $prop['links'] ) ) {
 202                         $result_array['links'] = $this->formatLinks( $p_result->getLinks() );
 203                 }
 204                 if ( isset( $prop['templates'] ) ) {
 205                         $result_array['templates'] = $this->formatLinks( $p_result->getTemplates() );
 206                 }
 207                 if ( isset( $prop['images'] ) ) {
 208                         $result_array['images'] = array_keys( $p_result->getImages() );
 209                 }
 210                 if ( isset( $prop['externallinks'] ) ) {
 211                         $result_array['externallinks'] = array_keys( $p_result->getExternalLinks() );
 212                 }
 213                 if ( isset( $prop['sections'] ) ) {
 214                         $result_array['sections'] = $p_result->getSections();
 215                 }
 216
 217                 if ( isset( $prop['displaytitle'] ) ) {
 218                         $result_array['displaytitle'] = $p_result->getDisplayTitle() ?
 219                                                         $p_result->getDisplayTitle() :
 220                                                         $titleObj->getPrefixedText();
 221                 }
 222
 223                 if ( isset( $prop['headitems'] ) || isset( $prop['headhtml'] ) ) {
 224                         $out = new OutputPage;
 225                         $out->addParserOutputNoText( $p_result );
 226                         $userSkin = $wgUser->getSkin();
 227                 }
 228
 229                 if ( isset( $prop['headitems'] ) ) {
 230                         $headItems = $this->formatHeadItems( $p_result->getHeadItems() );
 231
 232                         $userSkin->setupUserCss( $out );
 233                         $css = $this->formatCss( $out->buildCssLinksArray() );
 234
 235                         $scripts = array( $out->getHeadScripts( $userSkin ) );
 236
 237                         $result_array['headitems'] = array_merge( $headItems , $css, $scripts );
 238                 }
 239
 240                 if ( isset( $prop['headhtml'] ) ) {
 241                         $result_array['headhtml'] = array();
 242                         $result->setContent( $result_array['headhtml'], $out->headElement( $userSkin ) );
 243                 }
 244
 245                 if ( isset( $prop['iwlinks'] ) ) {
 246                         $result_array['iwlinks'] = $this->formatIWLinks( $p_result->getInterwikiLinks() );
 247                 }
 248
 249                 if ( !is_null( $oldid ) ) {
 250                         $result_array['revid'] = intval( $oldid );
 251                 }
 252
 253                 $result_mapping = array(
 254                         'redirects' => 'r',
 255                         'langlinks' => 'll',
 256                         'categories' => 'cl',
 257                         'links' => 'pl',
 258                         'templates' => 'tl',
 259                         'images' => 'img',
 260                         'externallinks' => 'el',
 261                         'iwlinks' => 'iw',
 262                         'sections' => 's',
 263                         'headitems' => 'hi',
 264                 );
 265                 $this->setIndexedTagNames( $result_array, $result_mapping );
 266                 $result->addValue( null, $this->getModuleName(), $result_array );
 267
 268                 if ( !is_null( $oldLang ) ) {
 269                         $wgLang = $oldLang; // Reset $wgLang to $oldLang
 270                 }
 271         }
 272
 273         private function getSectionText( $text, $what ) {
 274                 global $wgParser;
 275                 $text = $wgParser->getSection( $text, $this->section, false );
 276                 if ( $text === false ) {
 277                         $this->dieUsage( "There is no section {$this->section} in " . $what, 'nosuchsection' );
 278                 }
 279                 return $text;
 280         }
 281
 282         private function formatLangLinks( $links ) {
 283                 $result = array();
 284                 foreach ( $links as $link ) {
 285                         $entry = array();
 286                         $bits = explode( ':', $link, 2 );
 287                         $entry['lang'] = $bits[0];
 288                         $this->getResult()->setContent( $entry, $bits[1] );
 289                         $result[] = $entry;
 290                 }
 291                 return $result;
 292         }
 293
 294         private function formatCategoryLinks( $links ) {
 295                 $result = array();
 296                 foreach ( $links as $link => $sortkey ) {
 297                         $entry = array();
 298                         $entry['sortkey'] = $sortkey;
 299                         $this->getResult()->setContent( $entry, $link );
 300                         $result[] = $entry;
 301                 }
 302                 return $result;
 303         }
 304
 305         private function formatLinks( $links ) {
 306                 $result = array();
 307                 foreach ( $links as $ns => $nslinks ) {
 308                         foreach ( $nslinks as $title => $id ) {
 309                                 $entry = array();
 310                                 $entry['ns'] = $ns;
 311                                 $this->getResult()->setContent( $entry, Title::makeTitle( $ns, $title )->getFullText() );
 312                                 if ( $id != 0 ) {
 313                                         $entry['exists'] = '';
 314                                 }
 315                                 $result[] = $entry;
 316                         }
 317                 }
 318                 return $result;
 319         }
 320
 321         private function formatIWLinks( $iw ) {
 322                 $result = array();
 323                 foreach ( $iw as $prefix => $titles ) {
 324                         foreach ( $titles as $title => $id ) {
 325                                 $entry = array();
 326                                 $entry['prefix'] = $prefix;
 327
 328                                 $title = Title::newFromText( "{$prefix}:{$title}" );
 329                                 if ( $title ) {
 330                                         $entry['url'] = $title->getFullURL();
 331                                 }
 332
 333                                 $this->getResult()->setContent( $entry, $title->getFullText() );
 334                                 $result[] = $entry;
 335                         }
 336                 }
 337                 return $result;
 338         }
 339
 340         private function formatHeadItems( $headItems ) {
 341                 $result = array();
 342                 foreach ( $headItems as $tag => $content ) {
 343                         $entry = array();
 344                         $entry['tag'] = $tag;
 345                         $this->getResult()->setContent( $entry, $content );
 346                         $result[] = $entry;
 347                 }
 348                 return $result;
 349         }
 350
 351         private function formatCss( $css ) {
 352                 $result = array();
 353                 foreach ( $css as $file => $link ) {
 354                         $entry = array();
 355                         $entry['file'] = $file;
 356                         $this->getResult()->setContent( $entry, $link );
 357                         $result[] = $entry;
 358                 }
 359                 return $result;
 360         }
 361
 362         private function setIndexedTagNames( &$array, $mapping ) {
 363                 foreach ( $mapping as $key => $name ) {
 364                         if ( isset( $array[$key] ) ) {
 365                                 $this->getResult()->setIndexedTagName( $array[$key], $name );
 366                         }
 367                 }
 368         }
 369
 370         public function getAllowedParams() {
 371                 return array(
 372                         'title' => array(
 373                                 ApiBase::PARAM_DFLT => 'API',
 374                         ),
 375                         'text' => null,
 376                         'summary' => null,
 377                         'page' => null,
 378                         'pageid' => null,
 379                         'redirects' => false,
 380                         'oldid' => null,
 381                         'prop' => array(
 382                                 ApiBase::PARAM_DFLT => 'text|langlinks|categories|links|templates|images|externallinks|sections|revid|displaytitle',
 383                                 ApiBase::PARAM_ISMULTI => true,
 384                                 ApiBase::PARAM_TYPE => array(
 385                                         'text',
 386                                         'langlinks',
 387                                         'categories',
 388                                         'links',
 389                                         'templates',
 390                                         'images',
 391                                         'externallinks',
 392                                         'sections',
 393                                         'revid',
 394                                         'displaytitle',
 395                                         'headitems',
 396                                         'headhtml',
 397                                         'iwlinks',
 398                                 )
 399                         ),
 400                         'pst' => false,
 401                         'onlypst' => false,
 402                         'uselang' => null,
 403                         'section' => null,
 404                 );
 405         }
 406
 407         public function getParamDescription() {
 408                 $p = $this->getModulePrefix();
 409                 return array(
 410                         'text' => 'Wikitext to parse',
 411                         'summary' => 'Summary to parse',
 412                         'redirects' => "If the {$p}page parameter is set to a redirect, resolve it",
 413                         'title' => 'Title of page the text belongs to',
 414                         'page' => "Parse the content of this page. Cannot be used together with {$p}text and {$p}title",
 415                         'pageid' => "Parse the content of this page. Overrides {$p}page",
 416                         'oldid' => "Parse the content of this revision. Overrides {$p}page and {$p}pageid",
 417                         'prop' => array(
 418                                 'Which pieces of information to get',
 419                                 ' text           - Gives the parsed text of the wikitext',
 420                                 ' langlinks      - Gives the langlinks the parsed wikitext',
 421                                 ' categories     - Gives the categories of the parsed wikitext',
 422                                 ' links          - Gives the internal links in the parsed wikitext',
 423                                 ' templates      - Gives the templates in the parsed wikitext',
 424                                 ' images         - Gives the images in the parsed wikitext',
 425                                 ' externallinks  - Gives the external links in the parsed wikitext',
 426                                 ' sections       - Gives the sections in the parsed wikitext',
 427                                 ' revid          - Adds the revision id of the parsed page',
 428                                 ' displaytitle   - Adds the title of the parsed wikitext',
 429                                 ' headitems      - Gives items to put in the <head> of the page',
 430                                 ' headhtml       - Gives parsed <head> of the page',
 431                                 ' iwlinks        - Gives interwiki links in the parsed wikitext',
 432                                 'NOTE: Section tree is only generated if there are more than 4 sections, or if the __TOC__ keyword is present'
 433                         ),
 434                         'pst' => array(
 435                                 'Do a pre-save transform on the input before parsing it',
 436                                 'Ignored if page, pageid or oldid is used'
 437                         ),
 438                         'onlypst' => array(
 439                                 'Do a pre-save transform (PST) on the input, but don\'t parse it',
 440                                 'Returns the same wikitext, after a PST has been applied. Ignored if page, pageid or oldid is used'
 441                         ),
 442                         'uselang' => 'Which language to parse the request in',
 443                         'section' => 'Only retrieve the content of this section number',
 444                 );
 445         }
 446
 447         public function getDescription() {
 448                 return 'This module parses wikitext and returns parser output';
 449         }
 450
 451         public function getPossibleErrors() {
 452                 return array_merge( parent::getPossibleErrors(), array(
 453                         array( 'code' => 'params', 'info' => 'The page parameter cannot be used together with the text and title parameters' ),
 454                         array( 'code' => 'missingrev', 'info' => 'There is no revision ID oldid' ),
 455                         array( 'code' => 'permissiondenied', 'info' => 'You don\'t have permission to view deleted revisions' ),
 456                         array( 'code' => 'missingtitle', 'info' => 'The page you specified doesn\'t exist' ),
 457                         array( 'code' => 'nosuchsection', 'info' => 'There is no section sectionnumber in page' ),
 458                         array( 'nosuchpageid' ),
 459                 ) );
 460         }
 461
 462         protected function getExamples() {
 463                 return array(
 464                         'api.php?action=parse&text={{Project:Sandbox}}'
 465                 );
 466         }
 467
 468         public function getVersion() {
 469                 return __CLASS__ . ': $Id$';
 470         }
 471 }