includes/api/ApiParse.php

   1 <?php
   2 /**
   3  *
   4  *
   5  * Created on Dec 01, 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( "ApiBase.php" );
  30 }
  31
  32 /**
  33  * @ingroup API
  34  */
  35 class ApiParse extends ApiBase {
  36
  37         private $section;
  38
  39         public function __construct( $main, $action ) {
  40                 parent::__construct( $main, $action );
  41         }
  42
  43         public function execute() {
  44                 // The data is hot but user-dependent, like page views, so we set vary cookies
  45                 $this->getMain()->setCacheMode( 'anon-public-user-private' );
  46
  47                 // Get parameters
  48                 $params = $this->extractRequestParams();
  49                 $text = $params['text'];
  50                 $title = $params['title'];
  51                 $page = $params['page'];
  52                 $pageid = $params['pageid'];
  53                 $oldid = $params['oldid'];
  54
  55                 if ( !is_null( $page ) && ( !is_null( $text ) || $title != 'API' ) ) {
  56                         $this->dieUsage( 'The page parameter cannot be used together with the text and title parameters', 'params' );
  57                 }
  58                 $prop = array_flip( $params['prop'] );
  59
  60                 if ( isset( $params['section'] ) ) {
  61                         $this->section = $params['section'];
  62                 } else {
  63                         $this->section = false;
  64                 }
  65
  66                 // The parser needs $wgTitle to be set, apparently the
  67                 // $title parameter in Parser::parse isn't enough *sigh*
  68                 global $wgParser, $wgUser, $wgTitle, $wgEnableParserCache, $wgLang;
  69
  70                 // Currently unnecessary, code to act as a safeguard against any change in current behaviour of uselang breaks
  71                 $oldLang = null;
  72                 if ( isset( $params['uselang'] ) && $params['uselang'] != $wgLang->getCode() ) {
  73                         $oldLang = $wgLang; // Backup wgLang
  74                         $wgLang = Language::factory( $params['uselang'] );
  75                 }
  76
  77                 $popts = new ParserOptions();
  78                 $popts->setTidy( true );
  79                 $popts->enableLimitReport( !$params['disablepp'] );
  80                 $redirValues = null;
  81                 if ( !is_null( $oldid ) || !is_null( $pageid ) || !is_null( $page ) ) {
  82                         if ( !is_null( $oldid ) ) {
  83                                 // Don't use the parser cache
  84                                 $rev = Revision::newFromID( $oldid );
  85                                 if ( !$rev ) {
  86                                         $this->dieUsage( "There is no revision ID $oldid", 'missingrev' );
  87                                 }
  88                                 if ( !$rev->userCan( Revision::DELETED_TEXT ) ) {
  89                                         $this->dieUsage( "You don't have permission to view deleted revisions", 'permissiondenied' );
  90                                 }
  91
  92                                 $titleObj = $rev->getTitle();
  93
  94                                 $wgTitle = $titleObj;
  95
  96                                 // If for some reason the "oldid" is actually the current revision, it may be cached
  97                                 if ( $titleObj->getLatestRevID() === $oldid ) {
  98                                         $p_result = false;
  99                                         $pcache = ParserCache::singleton();
 100                                         if ( $wgEnableParserCache ) {
 101                                                 $p_result = $pcache->get( $titleObj, $popts );
 102                                         }
 103                                         if ( !$p_result ) {
 104                                                 $text = $rev->getText( Revision::FOR_THIS_USER );
 105                                                 $p_result = $wgParser->parse( $text, $titleObj, $popts );
 106
 107                                                 if ( $wgEnableParserCache ) {
 108                                                         $pcache->save( $p_result, $titleObj, $popts );
 109                                                 }
 110                                         }
 111                                 } else {
 112                                         $text = $rev->getText( Revision::FOR_THIS_USER );
 113
 114                                         $wgTitle = $titleObj;
 115
 116                                         if ( $this->section !== false ) {
 117                                                 $text = $this->getSectionText( $text, 'r' . $rev );
 118                                         }
 119
 120                                         $p_result = $wgParser->parse( $text, $titleObj, $popts );
 121                                 }
 122                         } else {
 123                                 if ( !is_null ( $pageid ) ) {
 124                                         $titleObj = Title::newFromID( $pageid );
 125
 126                                         if ( !$titleObj ) {
 127                                                 $this->dieUsageMsg( array( 'nosuchpageid', $pageid ) );
 128                                         }
 129                                 } else {
 130                                         if ( $params['redirects'] ) {
 131                                                 $req = new FauxRequest( array(
 132                                                         'action' => 'query',
 133                                                         'redirects' => '',
 134                                                         'titles' => $page
 135                                                 ) );
 136                                                 $main = new ApiMain( $req );
 137                                                 $main->execute();
 138                                                 $data = $main->getResultData();
 139                                                 $redirValues = @$data['query']['redirects'];
 140                                                 $to = $page;
 141                                                 foreach ( (array)$redirValues as $r ) {
 142                                                         $to = $r['to'];
 143                                                 }
 144                                         } else {
 145                                                 $to = $page;
 146                                         }
 147                                         $titleObj = Title::newFromText( $to );
 148                                         if ( !$titleObj ) {
 149                                                 $this->dieUsage( "The page you specified doesn't exist", 'missingtitle' );
 150                                         }
 151                                 }
 152                                 $wgTitle = $titleObj;
 153
 154                                 $articleObj = new Article( $titleObj );
 155                                 if ( isset( $prop['revid'] ) ) {
 156                                         $oldid = $articleObj->getRevIdFetched();
 157                                 }
 158
 159                                 if ( $this->section !== false ) {
 160                                         $text = $this->getSectionText( $text, !is_null ( $pageid ) ? 'page id ' . $pageid : $titleObj->getText() );
 161                                         $p_result = $wgParser->parse( $text, $titleObj, $popts );
 162                                 } else {
 163                                         // Try the parser cache first
 164                                         $p_result = false;
 165                                         $pcache = ParserCache::singleton();
 166                                         if ( $wgEnableParserCache ) {
 167                                                 $p_result = $pcache->get( $articleObj, $popts );
 168                                         }
 169                                         if ( !$p_result ) {
 170                                                 $p_result = $wgParser->parse( $articleObj->getContent(), $titleObj, $popts );
 171
 172                                                 if ( $wgEnableParserCache ) {
 173                                                         $pcache->save( $p_result, $articleObj, $popts );
 174                                                 }
 175                                         }
 176                                 }
 177                         }
 178                 } else {
 179                         $titleObj = Title::newFromText( $title );
 180                         if ( !$titleObj ) {
 181                                 $titleObj = Title::newFromText( 'API' );
 182                         }
 183                         $wgTitle = $titleObj;
 184
 185                         if ( $this->section !== false ) {
 186                                 $text = $this->getSectionText( $text, $titleObj->getText() );
 187                         }
 188
 189                         if ( $params['pst'] || $params['onlypst'] ) {
 190                                 $text = $wgParser->preSaveTransform( $text, $titleObj, $wgUser, $popts );
 191                         }
 192                         if ( $params['onlypst'] ) {
 193                                 // Build a result and bail out
 194                                 $result_array['text'] = array();
 195                                 $this->getResult()->setContent( $result_array['text'], $text );
 196                                 $this->getResult()->addValue( null, $this->getModuleName(), $result_array );
 197                                 return;
 198                         }
 199                         $p_result = $wgParser->parse( $text, $titleObj, $popts );
 200                 }
 201
 202                 // Return result
 203                 $result = $this->getResult();
 204                 $result_array = array();
 205                 if ( $params['redirects'] && !is_null( $redirValues ) ) {
 206                         $result_array['redirects'] = $redirValues;
 207                 }
 208
 209                 if ( isset( $prop['text'] ) ) {
 210                         $result_array['text'] = array();
 211                         $result->setContent( $result_array['text'], $p_result->getText() );
 212                 }
 213
 214                 if ( !is_null( $params['summary'] ) ) {
 215                         $result_array['parsedsummary'] = array();
 216                         $result->setContent( $result_array['parsedsummary'], $wgUser->getSkin()->formatComment( $params['summary'], $titleObj ) );
 217                 }
 218
 219                 if ( isset( $prop['langlinks'] ) ) {
 220                         $result_array['langlinks'] = $this->formatLangLinks( $p_result->getLanguageLinks() );
 221                 }
 222                 if ( isset( $prop['languageshtml'] ) ) {
 223                         $languagesHtml = $this->languagesHtml( $p_result->getLanguageLinks() );
 224                         $result_array['languageshtml'] = array();
 225                         $result->setContent( $result_array['languageshtml'], $languagesHtml );
 226                 }
 227                 if ( isset( $prop['categories'] ) ) {
 228                         $result_array['categories'] = $this->formatCategoryLinks( $p_result->getCategories() );
 229                 }
 230                 if ( isset( $prop['categorieshtml'] ) ) {
 231                         $categoriesHtml = $this->categoriesHtml( $p_result->getCategories() );
 232                         $result_array['categorieshtml'] = array();
 233                         $result->setContent( $result_array['categorieshtml'], $categoriesHtml );
 234                 }
 235                 if ( isset( $prop['links'] ) ) {
 236                         $result_array['links'] = $this->formatLinks( $p_result->getLinks() );
 237                 }
 238                 if ( isset( $prop['templates'] ) ) {
 239                         $result_array['templates'] = $this->formatLinks( $p_result->getTemplates() );
 240                 }
 241                 if ( isset( $prop['images'] ) ) {
 242                         $result_array['images'] = array_keys( $p_result->getImages() );
 243                 }
 244                 if ( isset( $prop['externallinks'] ) ) {
 245                         $result_array['externallinks'] = array_keys( $p_result->getExternalLinks() );
 246                 }
 247                 if ( isset( $prop['sections'] ) ) {
 248                         $result_array['sections'] = $p_result->getSections();
 249                 }
 250
 251                 if ( isset( $prop['displaytitle'] ) ) {
 252                         $result_array['displaytitle'] = $p_result->getDisplayTitle() ?
 253                                                         $p_result->getDisplayTitle() :
 254                                                         $titleObj->getPrefixedText();
 255                 }
 256
 257                 if ( isset( $prop['headitems'] ) || isset( $prop['headhtml'] ) ) {
 258                         $out = new OutputPage;
 259                         $out->addParserOutputNoText( $p_result );
 260                         $userSkin = $wgUser->getSkin();
 261                 }
 262
 263                 if ( isset( $prop['headitems'] ) ) {
 264                         $headItems = $this->formatHeadItems( $p_result->getHeadItems() );
 265
 266                         $userSkin->setupUserCss( $out );
 267                         $css = $this->formatCss( $out->buildCssLinksArray() );
 268
 269                         $scripts = array( $out->getHeadScripts( $userSkin ) );
 270
 271                         $result_array['headitems'] = array_merge( $headItems, $css, $scripts );
 272                 }
 273
 274                 if ( isset( $prop['headhtml'] ) ) {
 275                         $result_array['headhtml'] = array();
 276                         $result->setContent( $result_array['headhtml'], $out->headElement( $userSkin ) );
 277                 }
 278
 279                 if ( isset( $prop['iwlinks'] ) ) {
 280                         $result_array['iwlinks'] = $this->formatIWLinks( $p_result->getInterwikiLinks() );
 281                 }
 282
 283                 if ( !is_null( $oldid ) ) {
 284                         $result_array['revid'] = intval( $oldid );
 285                 }
 286
 287                 $result_mapping = array(
 288                         'redirects' => 'r',
 289                         'langlinks' => 'll',
 290                         'categories' => 'cl',
 291                         'links' => 'pl',
 292                         'templates' => 'tl',
 293                         'images' => 'img',
 294                         'externallinks' => 'el',
 295                         'iwlinks' => 'iw',
 296                         'sections' => 's',
 297                         'headitems' => 'hi',
 298                 );
 299                 $this->setIndexedTagNames( $result_array, $result_mapping );
 300                 $result->addValue( null, $this->getModuleName(), $result_array );
 301
 302                 if ( !is_null( $oldLang ) ) {
 303                         $wgLang = $oldLang; // Reset $wgLang to $oldLang
 304                 }
 305         }
 306
 307         private function getSectionText( $text, $what ) {
 308                 global $wgParser;
 309                 $text = $wgParser->getSection( $text, $this->section, false );
 310                 if ( $text === false ) {
 311                         $this->dieUsage( "There is no section {$this->section} in " . $what, 'nosuchsection' );
 312                 }
 313                 return $text;
 314         }
 315
 316         private function formatLangLinks( $links ) {
 317                 $result = array();
 318                 foreach ( $links as $link ) {
 319                         $entry = array();
 320                         $bits = explode( ':', $link, 2 );
 321                         $entry['lang'] = $bits[0];
 322                         $this->getResult()->setContent( $entry, $bits[1] );
 323                         $result[] = $entry;
 324                 }
 325                 return $result;
 326         }
 327
 328         private function formatCategoryLinks( $links ) {
 329                 $result = array();
 330                 foreach ( $links as $link => $sortkey ) {
 331                         $entry = array();
 332                         $entry['sortkey'] = $sortkey;
 333                         $this->getResult()->setContent( $entry, $link );
 334                         $result[] = $entry;
 335                 }
 336                 return $result;
 337         }
 338
 339         private function categoriesHtml( $categories ) {
 340                 global $wgOut, $wgUser;
 341                 $wgOut->addCategoryLinks( $categories );
 342                 $sk = $wgUser->getSkin();
 343                 return $sk->getCategories();
 344         }
 345
 346         private function languagesHtml( $languages ) {
 347                 global $wgOut, $wgUser;
 348                 $wgOut->setLanguageLinks( $languages );
 349                 $sk = $wgUser->getSkin();
 350                 return $sk->otherLanguages();
 351         }
 352
 353         private function formatLinks( $links ) {
 354                 $result = array();
 355                 foreach ( $links as $ns => $nslinks ) {
 356                         foreach ( $nslinks as $title => $id ) {
 357                                 $entry = array();
 358                                 $entry['ns'] = $ns;
 359                                 $this->getResult()->setContent( $entry, Title::makeTitle( $ns, $title )->getFullText() );
 360                                 if ( $id != 0 ) {
 361                                         $entry['exists'] = '';
 362                                 }
 363                                 $result[] = $entry;
 364                         }
 365                 }
 366                 return $result;
 367         }
 368
 369         private function formatIWLinks( $iw ) {
 370                 $result = array();
 371                 foreach ( $iw as $prefix => $titles ) {
 372                         foreach ( array_keys( $titles ) as $title ) {
 373                                 $entry = array();
 374                                 $entry['prefix'] = $prefix;
 375
 376                                 $title = Title::newFromText( "{$prefix}:{$title}" );
 377                                 if ( $title ) {
 378                                         $entry['url'] = $title->getFullURL();
 379                                 }
 380
 381                                 $this->getResult()->setContent( $entry, $title->getFullText() );
 382                                 $result[] = $entry;
 383                         }
 384                 }
 385                 return $result;
 386         }
 387
 388         private function formatHeadItems( $headItems ) {
 389                 $result = array();
 390                 foreach ( $headItems as $tag => $content ) {
 391                         $entry = array();
 392                         $entry['tag'] = $tag;
 393                         $this->getResult()->setContent( $entry, $content );
 394                         $result[] = $entry;
 395                 }
 396                 return $result;
 397         }
 398
 399         private function formatCss( $css ) {
 400                 $result = array();
 401                 foreach ( $css as $file => $link ) {
 402                         $entry = array();
 403                         $entry['file'] = $file;
 404                         $this->getResult()->setContent( $entry, $link );
 405                         $result[] = $entry;
 406                 }
 407                 return $result;
 408         }
 409
 410         private function setIndexedTagNames( &$array, $mapping ) {
 411                 foreach ( $mapping as $key => $name ) {
 412                         if ( isset( $array[$key] ) ) {
 413                                 $this->getResult()->setIndexedTagName( $array[$key], $name );
 414                         }
 415                 }
 416         }
 417
 418         public function getAllowedParams() {
 419                 return array(
 420                         'title' => array(
 421                                 ApiBase::PARAM_DFLT => 'API',
 422                         ),
 423                         'text' => null,
 424                         'summary' => null,
 425                         'page' => null,
 426                         'pageid' => null,
 427                         'redirects' => false,
 428                         'oldid' => null,
 429                         'prop' => array(
 430                                 ApiBase::PARAM_DFLT => 'text|langlinks|categories|links|templates|images|externallinks|sections|revid|displaytitle',
 431                                 ApiBase::PARAM_ISMULTI => true,
 432                                 ApiBase::PARAM_TYPE => array(
 433                                         'text',
 434                                         'langlinks',
 435                                         'languageshtml',
 436                                         'categories',
 437                                         'categorieshtml',
 438                                         'links',
 439                                         'templates',
 440                                         'images',
 441                                         'externallinks',
 442                                         'sections',
 443                                         'revid',
 444                                         'displaytitle',
 445                                         'headitems',
 446                                         'headhtml',
 447                                         'iwlinks',
 448                                 )
 449                         ),
 450                         'pst' => false,
 451                         'onlypst' => false,
 452                         'uselang' => null,
 453                         'section' => null,
 454                         'disablepp' => false,
 455                 );
 456         }
 457
 458         public function getParamDescription() {
 459                 $p = $this->getModulePrefix();
 460                 return array(
 461                         'text' => 'Wikitext to parse',
 462                         'summary' => 'Summary to parse',
 463                         'redirects' => "If the {$p}page parameter is set to a redirect, resolve it",
 464                         'title' => 'Title of page the text belongs to',
 465                         'page' => "Parse the content of this page. Cannot be used together with {$p}text and {$p}title",
 466                         'pageid' => "Parse the content of this page. Overrides {$p}page",
 467                         'oldid' => "Parse the content of this revision. Overrides {$p}page and {$p}pageid",
 468                         'prop' => array(
 469                                 'Which pieces of information to get',
 470                                 ' text           - Gives the parsed text of the wikitext',
 471                                 ' langlinks      - Gives the language links in the parsed wikitext',
 472                                 ' categories     - Gives the categories in the parsed wikitext',
 473                                 ' categorieshtml - Gives the HTML version of the categories',
 474                                 ' languageshtml  - Gives the HTML version of the language links',
 475                                 ' links          - Gives the internal links in the parsed wikitext',
 476                                 ' templates      - Gives the templates in the parsed wikitext',
 477                                 ' images         - Gives the images in the parsed wikitext',
 478                                 ' externallinks  - Gives the external links in the parsed wikitext',
 479                                 ' sections       - Gives the sections in the parsed wikitext',
 480                                 ' revid          - Adds the revision ID of the parsed page',
 481                                 ' displaytitle   - Adds the title of the parsed wikitext',
 482                                 ' headitems      - Gives items to put in the <head> of the page',
 483                                 ' headhtml       - Gives parsed <head> of the page',
 484                                 ' iwlinks        - Gives interwiki links in the parsed wikitext',
 485                                 'NOTE: Section tree is only generated if there are more than 4 sections, or if the __TOC__ keyword is present'
 486                         ),
 487                         'pst' => array(
 488                                 'Do a pre-save transform on the input before parsing it',
 489                                 'Ignored if page, pageid or oldid is used'
 490                         ),
 491                         'onlypst' => array(
 492                                 'Do a pre-save transform (PST) on the input, but don\'t parse it',
 493                                 'Returns the same wikitext, after a PST has been applied. Ignored if page, pageid or oldid is used'
 494                         ),
 495                         'uselang' => 'Which language to parse the request in',
 496                         'section' => 'Only retrieve the content of this section number',
 497                         'disablepp' => 'Disable the PP Report from the parser output',
 498                 );
 499         }
 500
 501         public function getDescription() {
 502                 return 'This module parses wikitext and returns parser output';
 503         }
 504
 505         public function getPossibleErrors() {
 506                 return array_merge( parent::getPossibleErrors(), array(
 507                         array( 'code' => 'params', 'info' => 'The page parameter cannot be used together with the text and title parameters' ),
 508                         array( 'code' => 'missingrev', 'info' => 'There is no revision ID oldid' ),
 509                         array( 'code' => 'permissiondenied', 'info' => 'You don\'t have permission to view deleted revisions' ),
 510                         array( 'code' => 'missingtitle', 'info' => 'The page you specified doesn\'t exist' ),
 511                         array( 'code' => 'nosuchsection', 'info' => 'There is no section sectionnumber in page' ),
 512                         array( 'nosuchpageid' ),
 513                 ) );
 514         }
 515
 516         protected function getExamples() {
 517                 return array(
 518                         'api.php?action=parse&text={{Project:Sandbox}}'
 519                 );
 520         }
 521
 522         public function getVersion() {
 523                 return __CLASS__ . ': $Id$';
 524         }
 525 }