includes/api/ApiParse.php

   1 <?php
   2 /**
   3  * Created on Dec 01, 2007
   4  *
   5  * Copyright © 2007 Yuri Astrakhan <Firstname><Lastname>@gmail.com
   6  *
   7  * This program is free software; you can redistribute it and/or modify
   8  * it under the terms of the GNU General Public License as published by
   9  * the Free Software Foundation; either version 2 of the License, or
  10  * (at your option) any later version.
  11  *
  12  * This program is distributed in the hope that it will be useful,
  13  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  14  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  15  * GNU General Public License for more details.
  16  *
  17  * You should have received a copy of the GNU General Public License along
  18  * with this program; if not, write to the Free Software Foundation, Inc.,
  19  * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
  20  * http://www.gnu.org/copyleft/gpl.html
  21  *
  22  * @file
  23  */
  24
  25 if ( !defined( 'MEDIAWIKI' ) ) {
  26         // Eclipse helper - will be ignored in production
  27         require_once( "ApiBase.php" );
  28 }
  29
  30 /**
  31  * @ingroup API
  32  */
  33 class ApiParse extends ApiBase {
  34         private $section, $text, $pstText = null;
  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, $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                 $popts->enableLimitReport( !$params['disablepp'] );
  77
  78                 $redirValues = null;
  79
  80                 if ( !is_null( $oldid ) || !is_null( $pageid ) || !is_null( $page ) ) {
  81                         if ( !is_null( $oldid ) ) {
  82                                 // Don't use the parser cache
  83                                 $rev = Revision::newFromID( $oldid );
  84                                 if ( !$rev ) {
  85                                         $this->dieUsage( "There is no revision ID $oldid", 'missingrev' );
  86                                 }
  87                                 if ( !$rev->userCan( Revision::DELETED_TEXT ) ) {
  88                                         $this->dieUsage( "You don't have permission to view deleted revisions", 'permissiondenied' );
  89                                 }
  90
  91                                 $titleObj = $rev->getTitle();
  92
  93                                 $wgTitle = $titleObj;
  94
  95                                 // If for some reason the "oldid" is actually the current revision, it may be cached
  96                                 if ( $titleObj->getLatestRevID() === intval( $oldid ) )  {
  97                                         $articleObj = new Article( $titleObj, 0 );
  98
  99                                         $p_result = $this->getParsedSectionOrText( $articleObj, $titleObj, $popts, $pageid,
 100                                                  isset( $prop['wikitext'] ) ) ;
 101                                 } else { // This is an old revision, so get the text differently
 102                                         $this->text = $rev->getText( Revision::FOR_THIS_USER );
 103
 104                                         $wgTitle = $titleObj;
 105
 106                                         if ( $this->section !== false ) {
 107                                                 $this->text = $this->getSectionText( $this->text, 'r' . $rev->getId() );
 108                                         }
 109
 110                                         $p_result = $wgParser->parse( $this->text, $titleObj, $popts );
 111                                 }
 112                         } else { // Not $oldid
 113                                 if ( !is_null ( $pageid ) ) {
 114                                         $titleObj = Title::newFromID( $pageid );
 115
 116                                         if ( !$titleObj ) {
 117                                                 $this->dieUsageMsg( array( 'nosuchpageid', $pageid ) );
 118                                         }
 119                                 } else { // $page
 120                                         if ( $params['redirects'] ) {
 121                                                 $req = new FauxRequest( array(
 122                                                         'action' => 'query',
 123                                                         'redirects' => '',
 124                                                         'titles' => $page
 125                                                 ) );
 126                                                 $main = new ApiMain( $req );
 127                                                 $main->execute();
 128                                                 $data = $main->getResultData();
 129                                                 $redirValues = isset( $data['query']['redirects'] )
 130                                                         ? $data['query']['redirects'] : array();
 131                                                 $to = $page;
 132                                                 foreach ( (array)$redirValues as $r ) {
 133                                                         $to = $r['to'];
 134                                                 }
 135                                         } else {
 136                                                 $to = $page;
 137                                         }
 138                                         $titleObj = Title::newFromText( $to );
 139                                         if ( !$titleObj || !$titleObj->exists() ) {
 140                                                 $this->dieUsage( "The page you specified doesn't exist", 'missingtitle' );
 141                                         }
 142                                 }
 143                                 $wgTitle = $titleObj;
 144
 145                                 $articleObj = new Article( $titleObj, 0 );
 146                                 if ( isset( $prop['revid'] ) ) {
 147                                         $oldid = $articleObj->getRevIdFetched();
 148                                 }
 149
 150                                 $p_result = $this->getParsedSectionOrText( $articleObj, $titleObj, $popts, $pageid,
 151                                          isset( $prop['wikitext'] ) ) ;
 152                         }
 153                 } else { // Not $oldid, $pageid, $page. Hence based on $text
 154
 155                         $this->text = $text;
 156                         $titleObj = Title::newFromText( $title );
 157                         if ( !$titleObj ) {
 158                                 $titleObj = Title::newFromText( 'API' );
 159                         }
 160                         $wgTitle = $titleObj;
 161
 162                         if ( $this->section !== false ) {
 163                                 $this->text = $this->getSectionText( $this->text, $titleObj->getText() );
 164                         }
 165
 166                         if ( $params['pst'] || $params['onlypst'] ) {
 167                                 $this->pstText = $wgParser->preSaveTransform( $this->text, $titleObj, $wgUser, $popts );
 168                         }
 169                         if ( $params['onlypst'] ) {
 170                                 // Build a result and bail out
 171                                 $result_array['text'] = array();
 172                                 $this->getResult()->setContent( $result_array['text'], $this->pstText );
 173                                 if ( isset( $prop['wikitext'] ) ) {
 174                                         $result_array['wikitext'] = array();
 175                                         $this->getResult()->setContent( $result_array['wikitext'], $this->text );
 176                                 }
 177                                 $this->getResult()->addValue( null, $this->getModuleName(), $result_array );
 178                                 return;
 179                         }
 180                         $p_result = $wgParser->parse( $params['pst'] ? $this->pstText : $this->text, $titleObj, $popts );
 181                 }
 182
 183                 // Return result
 184                 $result = $this->getResult();
 185                 $result_array = array();
 186
 187                 $result_array['title'] = $titleObj->getPrefixedText();
 188
 189                 if ( !is_null( $oldid ) ) {
 190                         $result_array['revid'] = intval( $oldid );
 191                 }
 192
 193                 if ( $params['redirects'] && !is_null( $redirValues ) ) {
 194                         $result_array['redirects'] = $redirValues;
 195                 }
 196
 197                 if ( isset( $prop['text'] ) ) {
 198                         $result_array['text'] = array();
 199                         $result->setContent( $result_array['text'], $p_result->getText() );
 200                 }
 201
 202                 if ( !is_null( $params['summary'] ) ) {
 203                         $result_array['parsedsummary'] = array();
 204                         $result->setContent( $result_array['parsedsummary'], $wgUser->getSkin()->formatComment( $params['summary'], $titleObj ) );
 205                 }
 206
 207                 if ( isset( $prop['langlinks'] ) ) {
 208                         $result_array['langlinks'] = $this->formatLangLinks( $p_result->getLanguageLinks() );
 209                 }
 210                 if ( isset( $prop['languageshtml'] ) ) {
 211                         $languagesHtml = $this->languagesHtml( $p_result->getLanguageLinks() );
 212                         $result_array['languageshtml'] = array();
 213                         $result->setContent( $result_array['languageshtml'], $languagesHtml );
 214                 }
 215                 if ( isset( $prop['categories'] ) ) {
 216                         $result_array['categories'] = $this->formatCategoryLinks( $p_result->getCategories() );
 217                 }
 218                 if ( isset( $prop['categorieshtml'] ) ) {
 219                         $categoriesHtml = $this->categoriesHtml( $p_result->getCategories() );
 220                         $result_array['categorieshtml'] = array();
 221                         $result->setContent( $result_array['categorieshtml'], $categoriesHtml );
 222                 }
 223                 if ( isset( $prop['links'] ) ) {
 224                         $result_array['links'] = $this->formatLinks( $p_result->getLinks() );
 225                 }
 226                 if ( isset( $prop['templates'] ) ) {
 227                         $result_array['templates'] = $this->formatLinks( $p_result->getTemplates() );
 228                 }
 229                 if ( isset( $prop['images'] ) ) {
 230                         $result_array['images'] = array_keys( $p_result->getImages() );
 231                 }
 232                 if ( isset( $prop['externallinks'] ) ) {
 233                         $result_array['externallinks'] = array_keys( $p_result->getExternalLinks() );
 234                 }
 235                 if ( isset( $prop['sections'] ) ) {
 236                         $result_array['sections'] = $p_result->getSections();
 237                 }
 238
 239                 if ( isset( $prop['displaytitle'] ) ) {
 240                         $result_array['displaytitle'] = $p_result->getDisplayTitle() ?
 241                                                         $p_result->getDisplayTitle() :
 242                                                         $titleObj->getPrefixedText();
 243                 }
 244
 245                 if ( isset( $prop['headitems'] ) || isset( $prop['headhtml'] ) ) {
 246                         $context = new RequestContext;
 247                         $context->output->addParserOutputNoText( $p_result );
 248
 249                         if ( isset( $prop['headitems'] ) ) {
 250                                 $headItems = $this->formatHeadItems( $p_result->getHeadItems() );
 251
 252                                 $context->skin->setupUserCss( $context->output );
 253                                 $css = $this->formatCss( $context->output->buildCssLinksArray() );
 254
 255                                 $scripts = array( $context->output->getHeadScripts( $context->skin ) );
 256
 257                                 $result_array['headitems'] = array_merge( $headItems, $css, $scripts );
 258                         }
 259
 260                         if ( isset( $prop['headhtml'] ) ) {
 261                                 $result_array['headhtml'] = array();
 262                                 $result->setContent( $result_array['headhtml'], $context->output->headElement( $context->skin ) );
 263                         }
 264                 }
 265
 266                 if ( isset( $prop['iwlinks'] ) ) {
 267                         $result_array['iwlinks'] = $this->formatIWLinks( $p_result->getInterwikiLinks() );
 268                 }
 269
 270                 if ( isset( $prop['wikitext'] ) ) {
 271                         $result_array['wikitext'] = array();
 272                         $result->setContent( $result_array['wikitext'], $this->text );
 273                         if ( !is_null( $this->pstText ) ) {
 274                                 $result_array['psttext'] = array();
 275                                 $result->setContent( $result_array['psttext'], $this->pstText );
 276                         }
 277                 }
 278
 279                 $result_mapping = array(
 280                         'redirects' => 'r',
 281                         'langlinks' => 'll',
 282                         'categories' => 'cl',
 283                         'links' => 'pl',
 284                         'templates' => 'tl',
 285                         'images' => 'img',
 286                         'externallinks' => 'el',
 287                         'iwlinks' => 'iw',
 288                         'sections' => 's',
 289                         'headitems' => 'hi',
 290                 );
 291                 $this->setIndexedTagNames( $result_array, $result_mapping );
 292                 $result->addValue( null, $this->getModuleName(), $result_array );
 293
 294                 if ( !is_null( $oldLang ) ) {
 295                         $wgLang = $oldLang; // Reset $wgLang to $oldLang
 296                 }
 297         }
 298
 299         /**
 300          * @param  $articleObj Article
 301          * @param  $titleObj Title
 302          * @param  $popts ParserOptions
 303          * @param  $pageId Int
 304          * @param  $getWikitext Bool
 305          * @return ParserOutput
 306          */
 307         private function getParsedSectionOrText( $articleObj, $titleObj, $popts, $pageId = null, $getWikitext = false ) {
 308                 if ( $this->section !== false ) {
 309                         global $wgParser;
 310
 311                         $this->text = $this->getSectionText( $articleObj->getRawText(), !is_null ( $pageId )
 312                                         ? 'page id ' . $pageId : $titleObj->getText() );
 313
 314                         return $wgParser->parse( $this->text, $titleObj, $popts );
 315                 } else {
 316                         // Try the parser cache first
 317                         $pout = $articleObj->getParserOutput();
 318                         if ( $getWikitext ) {
 319                                 $rev = Revision::newFromTitle( $titleObj );
 320                                 if ( $rev ) {
 321                                         $this->text = $rev->getText();
 322                                 }
 323                         }
 324                         return $pout;
 325                 }
 326         }
 327
 328         private function getSectionText( $text, $what ) {
 329                 global $wgParser;
 330                 $text = $wgParser->getSection( $text, $this->section, false );
 331                 if ( $text === false ) {
 332                         $this->dieUsage( "There is no section {$this->section} in " . $what, 'nosuchsection' );
 333                 }
 334                 return $text;
 335         }
 336
 337         private function formatLangLinks( $links ) {
 338                 $result = array();
 339                 foreach ( $links as $link ) {
 340                         $entry = array();
 341                         $bits = explode( ':', $link, 2 );
 342                         $title = Title::newFromText( $link );
 343
 344                         $entry['lang'] = $bits[0];
 345                         if ( $title ) {
 346                                 $entry['url'] = $title->getFullURL();
 347                         }
 348                         $this->getResult()->setContent( $entry, $bits[1] );
 349                         $result[] = $entry;
 350                 }
 351                 return $result;
 352         }
 353
 354         private function formatCategoryLinks( $links ) {
 355                 $result = array();
 356                 foreach ( $links as $link => $sortkey ) {
 357                         $entry = array();
 358                         $entry['sortkey'] = $sortkey;
 359                         $this->getResult()->setContent( $entry, $link );
 360                         $result[] = $entry;
 361                 }
 362                 return $result;
 363         }
 364
 365         private function categoriesHtml( $categories ) {
 366                 global $wgOut, $wgUser;
 367                 $wgOut->addCategoryLinks( $categories );
 368                 return $wgUser->getSkin()->getCategories();
 369         }
 370
 371         /**
 372          * @deprecated No modern skin generates langlinks this way, please use langlinks data to generate your own html
 373          */
 374         private function languagesHtml( $languages ) {
 375                 global $wgContLang, $wgHideInterlanguageLinks;
 376
 377                 if ( $wgHideInterlanguageLinks || count( $languages ) == 0 ) {
 378                         return '';
 379                 }
 380
 381                 $s = htmlspecialchars( wfMsg( 'otherlanguages' ) . wfMsg( 'colon-separator' ) );
 382
 383                 $langs = array();
 384                 foreach ( $languages as $l ) {
 385                         $nt = Title::newFromText( $l );
 386                         $text = $wgContLang->getLanguageName( $nt->getInterwiki() );
 387
 388                         $langs[] = Html::element( 'a',
 389                                 array( 'href' => $nt->getFullURL(), 'title' => $nt->getText(), 'class' => "external" ),
 390                                 $text == '' ? $l : $text );
 391                 }
 392
 393                 $s .= implode( htmlspecialchars( wfMsgExt( 'pipe-separator', 'escapenoentities' ) ), $langs );
 394
 395                 if ( $wgContLang->isRTL() ) {
 396                         $s = Html::rawElement( 'span', array( 'dir' => "LTR" ), $s );
 397                 }
 398
 399                 return $s;
 400         }
 401
 402         private function formatLinks( $links ) {
 403                 $result = array();
 404                 foreach ( $links as $ns => $nslinks ) {
 405                         foreach ( $nslinks as $title => $id ) {
 406                                 $entry = array();
 407                                 $entry['ns'] = $ns;
 408                                 $this->getResult()->setContent( $entry, Title::makeTitle( $ns, $title )->getFullText() );
 409                                 if ( $id != 0 ) {
 410                                         $entry['exists'] = '';
 411                                 }
 412                                 $result[] = $entry;
 413                         }
 414                 }
 415                 return $result;
 416         }
 417
 418         private function formatIWLinks( $iw ) {
 419                 $result = array();
 420                 foreach ( $iw as $prefix => $titles ) {
 421                         foreach ( array_keys( $titles ) as $title ) {
 422                                 $entry = array();
 423                                 $entry['prefix'] = $prefix;
 424
 425                                 $title = Title::newFromText( "{$prefix}:{$title}" );
 426                                 if ( $title ) {
 427                                         $entry['url'] = $title->getFullURL();
 428                                 }
 429
 430                                 $this->getResult()->setContent( $entry, $title->getFullText() );
 431                                 $result[] = $entry;
 432                         }
 433                 }
 434                 return $result;
 435         }
 436
 437         private function formatHeadItems( $headItems ) {
 438                 $result = array();
 439                 foreach ( $headItems as $tag => $content ) {
 440                         $entry = array();
 441                         $entry['tag'] = $tag;
 442                         $this->getResult()->setContent( $entry, $content );
 443                         $result[] = $entry;
 444                 }
 445                 return $result;
 446         }
 447
 448         private function formatCss( $css ) {
 449                 $result = array();
 450                 foreach ( $css as $file => $link ) {
 451                         $entry = array();
 452                         $entry['file'] = $file;
 453                         $this->getResult()->setContent( $entry, $link );
 454                         $result[] = $entry;
 455                 }
 456                 return $result;
 457         }
 458
 459         private function setIndexedTagNames( &$array, $mapping ) {
 460                 foreach ( $mapping as $key => $name ) {
 461                         if ( isset( $array[$key] ) ) {
 462                                 $this->getResult()->setIndexedTagName( $array[$key], $name );
 463                         }
 464                 }
 465         }
 466
 467         public function getAllowedParams() {
 468                 return array(
 469                         'title' => array(
 470                                 ApiBase::PARAM_DFLT => 'API',
 471                         ),
 472                         'text' => null,
 473                         'summary' => null,
 474                         'page' => null,
 475                         'pageid' => array(
 476                                 ApiBase::PARAM_TYPE => 'integer',
 477                         ),
 478                         'redirects' => false,
 479                         'oldid' => array(
 480                                 ApiBase::PARAM_TYPE => 'integer',
 481                         ),
 482                         'prop' => array(
 483                                 ApiBase::PARAM_DFLT => 'text|langlinks|categories|links|templates|images|externallinks|sections|revid|displaytitle',
 484                                 ApiBase::PARAM_ISMULTI => true,
 485                                 ApiBase::PARAM_TYPE => array(
 486                                         'text',
 487                                         'langlinks',
 488                                         'languageshtml',
 489                                         'categories',
 490                                         'categorieshtml',
 491                                         'links',
 492                                         'templates',
 493                                         'images',
 494                                         'externallinks',
 495                                         'sections',
 496                                         'revid',
 497                                         'displaytitle',
 498                                         'headitems',
 499                                         'headhtml',
 500                                         'iwlinks',
 501                                         'wikitext',
 502                                 )
 503                         ),
 504                         'pst' => false,
 505                         'onlypst' => false,
 506                         'uselang' => null,
 507                         'section' => null,
 508                         'disablepp' => false,
 509                 );
 510         }
 511
 512         public function getParamDescription() {
 513                 $p = $this->getModulePrefix();
 514                 return array(
 515                         'text' => 'Wikitext to parse',
 516                         'summary' => 'Summary to parse',
 517                         'redirects' => "If the {$p}page parameter is set to a redirect, resolve it",
 518                         'title' => 'Title of page the text belongs to',
 519                         'page' => "Parse the content of this page. Cannot be used together with {$p}text and {$p}title",
 520                         'pageid' => "Parse the content of this page. Overrides {$p}page",
 521                         'oldid' => "Parse the content of this revision. Overrides {$p}page and {$p}pageid",
 522                         'prop' => array(
 523                                 'Which pieces of information to get',
 524                                 ' text           - Gives the parsed text of the wikitext',
 525                                 ' langlinks      - Gives the language links in the parsed wikitext',
 526                                 ' categories     - Gives the categories in the parsed wikitext',
 527                                 ' categorieshtml - Gives the HTML version of the categories',
 528                                 ' languageshtml  - Gives the HTML version of the language links',
 529                                 ' links          - Gives the internal links in the parsed wikitext',
 530                                 ' templates      - Gives the templates in the parsed wikitext',
 531                                 ' images         - Gives the images in the parsed wikitext',
 532                                 ' externallinks  - Gives the external links in the parsed wikitext',
 533                                 ' sections       - Gives the sections in the parsed wikitext',
 534                                 ' revid          - Adds the revision ID of the parsed page',
 535                                 ' displaytitle   - Adds the title of the parsed wikitext',
 536                                 ' headitems      - Gives items to put in the <head> of the page',
 537                                 ' headhtml       - Gives parsed <head> of the page',
 538                                 ' iwlinks        - Gives interwiki links in the parsed wikitext',
 539                                 ' wikitext       - Gives the original wikitext that was parsed',
 540                         ),
 541                         'pst' => array(
 542                                 'Do a pre-save transform on the input before parsing it',
 543                                 'Ignored if page, pageid or oldid is used'
 544                         ),
 545                         'onlypst' => array(
 546                                 'Do a pre-save transform (PST) on the input, but don\'t parse it',
 547                                 'Returns the same wikitext, after a PST has been applied. Ignored if page, pageid or oldid is used'
 548                         ),
 549                         'uselang' => 'Which language to parse the request in',
 550                         'section' => 'Only retrieve the content of this section number',
 551                         'disablepp' => 'Disable the PP Report from the parser output',
 552                 );
 553         }
 554
 555         public function getDescription() {
 556                 return 'Parses wikitext and returns parser output';
 557         }
 558
 559         public function getPossibleErrors() {
 560                 return array_merge( parent::getPossibleErrors(), array(
 561                         array( 'code' => 'params', 'info' => 'The page parameter cannot be used together with the text and title parameters' ),
 562                         array( 'code' => 'missingrev', 'info' => 'There is no revision ID oldid' ),
 563                         array( 'code' => 'permissiondenied', 'info' => 'You don\'t have permission to view deleted revisions' ),
 564                         array( 'code' => 'missingtitle', 'info' => 'The page you specified doesn\'t exist' ),
 565                         array( 'code' => 'nosuchsection', 'info' => 'There is no section sectionnumber in page' ),
 566                         array( 'nosuchpageid' ),
 567                 ) );
 568         }
 569
 570         protected function getExamples() {
 571                 return array(
 572                         'api.php?action=parse&text={{Project:Sandbox}}'
 573                 );
 574         }
 575
 576         public function getVersion() {
 577                 return __CLASS__ . ': $Id$';
 578         }
 579 }