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                 // Return result
  81                 $result = $this->getResult();
  82
  83                 if ( !is_null( $oldid ) || !is_null( $pageid ) || !is_null( $page ) ) {
  84                         if ( !is_null( $oldid ) ) {
  85                                 // Don't use the parser cache
  86                                 $rev = Revision::newFromID( $oldid );
  87                                 if ( !$rev ) {
  88                                         $this->dieUsage( "There is no revision ID $oldid", 'missingrev' );
  89                                 }
  90                                 if ( !$rev->userCan( Revision::DELETED_TEXT ) ) {
  91                                         $this->dieUsage( "You don't have permission to view deleted revisions", 'permissiondenied' );
  92                                 }
  93
  94                                 $titleObj = $rev->getTitle();
  95
  96                                 $wgTitle = $titleObj;
  97
  98                                 // If for some reason the "oldid" is actually the current revision, it may be cached
  99                                 if ( $titleObj->getLatestRevID() === intval( $oldid ) )  {
 100                                         $articleObj = new Article( $titleObj, 0 );
 101
 102                                         $p_result = $this->getParsedSectionOrText( $articleObj, $titleObj, $popts, $pageid,
 103                                                  isset( $prop['wikitext'] ) ) ;
 104                                 } else { // This is an old revision, so get the text differently
 105                                         $this->text = $rev->getText( Revision::FOR_THIS_USER );
 106
 107                                         $wgTitle = $titleObj;
 108
 109                                         if ( $this->section !== false ) {
 110                                                 $this->text = $this->getSectionText( $this->text, 'r' . $rev->getId() );
 111                                         }
 112
 113                                         $p_result = $wgParser->parse( $this->text, $titleObj, $popts );
 114                                 }
 115                         } else { // Not $oldid
 116                                 if ( $params['redirects'] ) {
 117                                         $reqParams = array(
 118                                                 'action' => 'query',
 119                                                 'redirects' => '',
 120                                         );
 121                                         if ( !is_null ( $pageid ) ) {
 122                                                 $reqParams['pageids'] = $pageid;
 123                                         } else { // $page
 124                                                 $reqParams['titles'] = $page;
 125                                         }
 126                                         $req = new FauxRequest( $reqParams );
 127                                         $main = new ApiMain( $req );
 128                                         $main->execute();
 129                                         $data = $main->getResultData();
 130                                         $redirValues = isset( $data['query']['redirects'] )
 131                                                 ? $data['query']['redirects']
 132                                                 : array();
 133                                         $to = $page;
 134                                         foreach ( (array)$redirValues as $r ) {
 135                                                 $to = $r['to'];
 136                                         }
 137                                         $titleObj = Title::newFromText( $to );
 138                                 } else {
 139                                         if ( !is_null ( $pageid ) ) {
 140                                                 $reqParams['pageids'] = $pageid;
 141                                                 $titleObj = Title::newFromID( $pageid );
 142                                         } else { // $page
 143                                                 $to = $page;
 144                                                 $titleObj = Title::newFromText( $to );
 145                                         }
 146                                 }
 147                                 if ( !is_null ( $pageid ) ) {
 148                                         if ( !$titleObj ) {
 149                                                 // Still throw nosuchpageid error if pageid was provided
 150                                                 $this->dieUsageMsg( array( 'nosuchpageid', $pageid ) );
 151                                         }
 152                                 } elseif ( !$titleObj || !$titleObj->exists() ) {
 153                                         $this->dieUsage( "The page you specified doesn't exist", 'missingtitle' );
 154                                 }
 155                                 $wgTitle = $titleObj;
 156
 157                                 $articleObj = new Article( $titleObj, 0 );
 158                                 if ( isset( $prop['revid'] ) ) {
 159                                         $oldid = $articleObj->getRevIdFetched();
 160                                 }
 161
 162                                 $p_result = $this->getParsedSectionOrText( $articleObj, $titleObj, $popts, $pageid,
 163                                          isset( $prop['wikitext'] ) ) ;
 164                         }
 165                 } else { // Not $oldid, $pageid, $page. Hence based on $text
 166
 167                         $this->text = $text;
 168                         $titleObj = Title::newFromText( $title );
 169                         if ( !$titleObj ) {
 170                                 $this->dieUsageMsg( array( 'invalidtitle', $title ) );
 171                         }
 172                         $wgTitle = $titleObj;
 173
 174                         if ( $this->section !== false ) {
 175                                 $this->text = $this->getSectionText( $this->text, $titleObj->getText() );
 176                         }
 177
 178                         if ( $params['pst'] || $params['onlypst'] ) {
 179                                 $this->pstText = $wgParser->preSaveTransform( $this->text, $titleObj, $wgUser, $popts );
 180                         }
 181                         if ( $params['onlypst'] ) {
 182                                 // Build a result and bail out
 183                                 $result_array['text'] = array();
 184                                 $result->setContent( $result_array['text'], $this->pstText );
 185                                 if ( isset( $prop['wikitext'] ) ) {
 186                                         $result_array['wikitext'] = array();
 187                                         $result->setContent( $result_array['wikitext'], $this->text );
 188                                 }
 189                                 $result->addValue( null, $this->getModuleName(), $result_array );
 190                                 return;
 191                         }
 192                         $p_result = $wgParser->parse( $params['pst'] ? $this->pstText : $this->text, $titleObj, $popts );
 193                 }
 194
 195                 $result_array = array();
 196
 197                 $result_array['title'] = $titleObj->getPrefixedText();
 198
 199                 if ( !is_null( $oldid ) ) {
 200                         $result_array['revid'] = intval( $oldid );
 201                 }
 202
 203                 if ( $params['redirects'] && !is_null( $redirValues ) ) {
 204                         $result_array['redirects'] = $redirValues;
 205                 }
 206
 207                 if ( isset( $prop['text'] ) ) {
 208                         $result_array['text'] = array();
 209                         $result->setContent( $result_array['text'], $p_result->getText() );
 210                 }
 211
 212                 if ( !is_null( $params['summary'] ) ) {
 213                         $result_array['parsedsummary'] = array();
 214                         $result->setContent( $result_array['parsedsummary'], $wgUser->getSkin()->formatComment( $params['summary'], $titleObj ) );
 215                 }
 216
 217                 if ( isset( $prop['langlinks'] ) ) {
 218                         $result_array['langlinks'] = $this->formatLangLinks( $p_result->getLanguageLinks() );
 219                 }
 220                 if ( isset( $prop['languageshtml'] ) ) {
 221                         $languagesHtml = $this->languagesHtml( $p_result->getLanguageLinks() );
 222                         $result_array['languageshtml'] = array();
 223                         $result->setContent( $result_array['languageshtml'], $languagesHtml );
 224                 }
 225                 if ( isset( $prop['categories'] ) ) {
 226                         $result_array['categories'] = $this->formatCategoryLinks( $p_result->getCategories() );
 227                 }
 228                 if ( isset( $prop['categorieshtml'] ) ) {
 229                         $categoriesHtml = $this->categoriesHtml( $p_result->getCategories() );
 230                         $result_array['categorieshtml'] = array();
 231                         $result->setContent( $result_array['categorieshtml'], $categoriesHtml );
 232                 }
 233                 if ( isset( $prop['links'] ) ) {
 234                         $result_array['links'] = $this->formatLinks( $p_result->getLinks() );
 235                 }
 236                 if ( isset( $prop['templates'] ) ) {
 237                         $result_array['templates'] = $this->formatLinks( $p_result->getTemplates() );
 238                 }
 239                 if ( isset( $prop['images'] ) ) {
 240                         $result_array['images'] = array_keys( $p_result->getImages() );
 241                 }
 242                 if ( isset( $prop['externallinks'] ) ) {
 243                         $result_array['externallinks'] = array_keys( $p_result->getExternalLinks() );
 244                 }
 245                 if ( isset( $prop['sections'] ) ) {
 246                         $result_array['sections'] = $p_result->getSections();
 247                 }
 248
 249                 if ( isset( $prop['displaytitle'] ) ) {
 250                         $result_array['displaytitle'] = $p_result->getDisplayTitle() ?
 251                                                         $p_result->getDisplayTitle() :
 252                                                         $titleObj->getPrefixedText();
 253                 }
 254
 255                 if ( isset( $prop['headitems'] ) || isset( $prop['headhtml'] ) ) {
 256                         $context = new RequestContext;
 257                         $context->getOutput()->addParserOutputNoText( $p_result );
 258
 259                         if ( isset( $prop['headitems'] ) ) {
 260                                 $headItems = $this->formatHeadItems( $p_result->getHeadItems() );
 261
 262                                 $context->getSkin()->setupUserCss( $context->getOutput() );
 263                                 $css = $this->formatCss( $context->getOutput()->buildCssLinksArray() );
 264
 265                                 $scripts = array( $context->getOutput()->getHeadScripts( $context->getSkin() ) );
 266
 267                                 $result_array['headitems'] = array_merge( $headItems, $css, $scripts );
 268                         }
 269
 270                         if ( isset( $prop['headhtml'] ) ) {
 271                                 $result_array['headhtml'] = array();
 272                                 $result->setContent( $result_array['headhtml'], $context->getOutput()->headElement( $context->getSkin() ) );
 273                         }
 274                 }
 275
 276                 if ( isset( $prop['iwlinks'] ) ) {
 277                         $result_array['iwlinks'] = $this->formatIWLinks( $p_result->getInterwikiLinks() );
 278                 }
 279
 280                 if ( isset( $prop['wikitext'] ) ) {
 281                         $result_array['wikitext'] = array();
 282                         $result->setContent( $result_array['wikitext'], $this->text );
 283                         if ( !is_null( $this->pstText ) ) {
 284                                 $result_array['psttext'] = array();
 285                                 $result->setContent( $result_array['psttext'], $this->pstText );
 286                         }
 287                 }
 288
 289                 $result_mapping = array(
 290                         'redirects' => 'r',
 291                         'langlinks' => 'll',
 292                         'categories' => 'cl',
 293                         'links' => 'pl',
 294                         'templates' => 'tl',
 295                         'images' => 'img',
 296                         'externallinks' => 'el',
 297                         'iwlinks' => 'iw',
 298                         'sections' => 's',
 299                         'headitems' => 'hi',
 300                 );
 301                 $this->setIndexedTagNames( $result_array, $result_mapping );
 302                 $result->addValue( null, $this->getModuleName(), $result_array );
 303
 304                 if ( !is_null( $oldLang ) ) {
 305                         $wgLang = $oldLang; // Reset $wgLang to $oldLang
 306                 }
 307         }
 308
 309         /**
 310          * @param $articleObj Article
 311          * @param $titleObj Title
 312          * @param $popts ParserOptions
 313          * @param $pageId Int
 314          * @param $getWikitext Bool
 315          * @return ParserOutput
 316          */
 317         private function getParsedSectionOrText( $articleObj, $titleObj, $popts, $pageId = null, $getWikitext = false ) {
 318                 if ( $this->section !== false ) {
 319                         global $wgParser;
 320
 321                         $this->text = $this->getSectionText( $articleObj->getRawText(), !is_null ( $pageId )
 322                                         ? 'page id ' . $pageId : $titleObj->getText() );
 323
 324                         return $wgParser->parse( $this->text, $titleObj, $popts );
 325                 } else {
 326                         // Try the parser cache first
 327                         $pout = $articleObj->getParserOutput();
 328                         if ( $getWikitext ) {
 329                                 $rev = Revision::newFromTitle( $titleObj );
 330                                 if ( $rev ) {
 331                                         $this->text = $rev->getText();
 332                                 }
 333                         }
 334                         return $pout;
 335                 }
 336         }
 337
 338         private function getSectionText( $text, $what ) {
 339                 global $wgParser;
 340                 $text = $wgParser->getSection( $text, $this->section, false );
 341                 if ( $text === false ) {
 342                         $this->dieUsage( "There is no section {$this->section} in " . $what, 'nosuchsection' );
 343                 }
 344                 return $text;
 345         }
 346
 347         private function formatLangLinks( $links ) {
 348                 $result = array();
 349                 foreach ( $links as $link ) {
 350                         $entry = array();
 351                         $bits = explode( ':', $link, 2 );
 352                         $title = Title::newFromText( $link );
 353
 354                         $entry['lang'] = $bits[0];
 355                         if ( $title ) {
 356                                 $entry['url'] = $title->getFullURL();
 357                         }
 358                         $this->getResult()->setContent( $entry, $bits[1] );
 359                         $result[] = $entry;
 360                 }
 361                 return $result;
 362         }
 363
 364         private function formatCategoryLinks( $links ) {
 365                 $result = array();
 366                 foreach ( $links as $link => $sortkey ) {
 367                         $entry = array();
 368                         $entry['sortkey'] = $sortkey;
 369                         $this->getResult()->setContent( $entry, $link );
 370                         $result[] = $entry;
 371                 }
 372                 return $result;
 373         }
 374
 375         private function categoriesHtml( $categories ) {
 376                 $context = $this->createContext();
 377                 $context->getOutput()->addCategoryLinks( $categories );
 378                 return $context->getSkin()->getCategories();
 379         }
 380
 381         /**
 382          * @deprecated since 1.18 No modern skin generates language links this way, please use language links
 383          *                        data to generate your own HTML.
 384          */
 385         private function languagesHtml( $languages ) {
 386                 global $wgContLang, $wgHideInterlanguageLinks;
 387
 388                 if ( $wgHideInterlanguageLinks || count( $languages ) == 0 ) {
 389                         return '';
 390                 }
 391
 392                 $s = htmlspecialchars( wfMsg( 'otherlanguages' ) . wfMsg( 'colon-separator' ) );
 393
 394                 $langs = array();
 395                 foreach ( $languages as $l ) {
 396                         $nt = Title::newFromText( $l );
 397                         $text = $wgContLang->getLanguageName( $nt->getInterwiki() );
 398
 399                         $langs[] = Html::element( 'a',
 400                                 array( 'href' => $nt->getFullURL(), 'title' => $nt->getText(), 'class' => "external" ),
 401                                 $text == '' ? $l : $text );
 402                 }
 403
 404                 $s .= implode( htmlspecialchars( wfMsgExt( 'pipe-separator', 'escapenoentities' ) ), $langs );
 405
 406                 if ( $wgContLang->isRTL() ) {
 407                         $s = Html::rawElement( 'span', array( 'dir' => "LTR" ), $s );
 408                 }
 409
 410                 return $s;
 411         }
 412
 413         private function formatLinks( $links ) {
 414                 $result = array();
 415                 foreach ( $links as $ns => $nslinks ) {
 416                         foreach ( $nslinks as $title => $id ) {
 417                                 $entry = array();
 418                                 $entry['ns'] = $ns;
 419                                 $this->getResult()->setContent( $entry, Title::makeTitle( $ns, $title )->getFullText() );
 420                                 if ( $id != 0 ) {
 421                                         $entry['exists'] = '';
 422                                 }
 423                                 $result[] = $entry;
 424                         }
 425                 }
 426                 return $result;
 427         }
 428
 429         private function formatIWLinks( $iw ) {
 430                 $result = array();
 431                 foreach ( $iw as $prefix => $titles ) {
 432                         foreach ( array_keys( $titles ) as $title ) {
 433                                 $entry = array();
 434                                 $entry['prefix'] = $prefix;
 435
 436                                 $title = Title::newFromText( "{$prefix}:{$title}" );
 437                                 if ( $title ) {
 438                                         $entry['url'] = $title->getFullURL();
 439                                 }
 440
 441                                 $this->getResult()->setContent( $entry, $title->getFullText() );
 442                                 $result[] = $entry;
 443                         }
 444                 }
 445                 return $result;
 446         }
 447
 448         private function formatHeadItems( $headItems ) {
 449                 $result = array();
 450                 foreach ( $headItems as $tag => $content ) {
 451                         $entry = array();
 452                         $entry['tag'] = $tag;
 453                         $this->getResult()->setContent( $entry, $content );
 454                         $result[] = $entry;
 455                 }
 456                 return $result;
 457         }
 458
 459         private function formatCss( $css ) {
 460                 $result = array();
 461                 foreach ( $css as $file => $link ) {
 462                         $entry = array();
 463                         $entry['file'] = $file;
 464                         $this->getResult()->setContent( $entry, $link );
 465                         $result[] = $entry;
 466                 }
 467                 return $result;
 468         }
 469
 470         private function setIndexedTagNames( &$array, $mapping ) {
 471                 foreach ( $mapping as $key => $name ) {
 472                         if ( isset( $array[$key] ) ) {
 473                                 $this->getResult()->setIndexedTagName( $array[$key], $name );
 474                         }
 475                 }
 476         }
 477
 478         public function getAllowedParams() {
 479                 return array(
 480                         'title' => array(
 481                                 ApiBase::PARAM_DFLT => 'API',
 482                         ),
 483                         'text' => null,
 484                         'summary' => null,
 485                         'page' => null,
 486                         'pageid' => array(
 487                                 ApiBase::PARAM_TYPE => 'integer',
 488                         ),
 489                         'redirects' => false,
 490                         'oldid' => array(
 491                                 ApiBase::PARAM_TYPE => 'integer',
 492                         ),
 493                         'prop' => array(
 494                                 ApiBase::PARAM_DFLT => 'text|langlinks|categories|links|templates|images|externallinks|sections|revid|displaytitle',
 495                                 ApiBase::PARAM_ISMULTI => true,
 496                                 ApiBase::PARAM_TYPE => array(
 497                                         'text',
 498                                         'langlinks',
 499                                         'languageshtml',
 500                                         'categories',
 501                                         'categorieshtml',
 502                                         'links',
 503                                         'templates',
 504                                         'images',
 505                                         'externallinks',
 506                                         'sections',
 507                                         'revid',
 508                                         'displaytitle',
 509                                         'headitems',
 510                                         'headhtml',
 511                                         'iwlinks',
 512                                         'wikitext',
 513                                 )
 514                         ),
 515                         'pst' => false,
 516                         'onlypst' => false,
 517                         'uselang' => null,
 518                         'section' => null,
 519                         'disablepp' => false,
 520                 );
 521         }
 522
 523         public function getParamDescription() {
 524                 $p = $this->getModulePrefix();
 525                 return array(
 526                         'text' => 'Wikitext to parse',
 527                         'summary' => 'Summary to parse',
 528                         'redirects' => "If the {$p}page or the {$p}pageid parameter is set to a redirect, resolve it",
 529                         'title' => 'Title of page the text belongs to',
 530                         'page' => "Parse the content of this page. Cannot be used together with {$p}text and {$p}title",
 531                         'pageid' => "Parse the content of this page. Overrides {$p}page",
 532                         'oldid' => "Parse the content of this revision. Overrides {$p}page and {$p}pageid",
 533                         'prop' => array(
 534                                 'Which pieces of information to get',
 535                                 ' text           - Gives the parsed text of the wikitext',
 536                                 ' langlinks      - Gives the language links in the parsed wikitext',
 537                                 ' categories     - Gives the categories in the parsed wikitext',
 538                                 ' categorieshtml - Gives the HTML version of the categories',
 539                                 ' languageshtml  - Gives the HTML version of the language links',
 540                                 ' links          - Gives the internal links in the parsed wikitext',
 541                                 ' templates      - Gives the templates in the parsed wikitext',
 542                                 ' images         - Gives the images in the parsed wikitext',
 543                                 ' externallinks  - Gives the external links in the parsed wikitext',
 544                                 ' sections       - Gives the sections in the parsed wikitext',
 545                                 ' revid          - Adds the revision ID of the parsed page',
 546                                 ' displaytitle   - Adds the title of the parsed wikitext',
 547                                 ' headitems      - Gives items to put in the <head> of the page',
 548                                 ' headhtml       - Gives parsed <head> of the page',
 549                                 ' iwlinks        - Gives interwiki links in the parsed wikitext',
 550                                 ' wikitext       - Gives the original wikitext that was parsed',
 551                         ),
 552                         'pst' => array(
 553                                 'Do a pre-save transform on the input before parsing it',
 554                                 'Ignored if page, pageid or oldid is used'
 555                         ),
 556                         'onlypst' => array(
 557                                 'Do a pre-save transform (PST) on the input, but don\'t parse it',
 558                                 'Returns the same wikitext, after a PST has been applied. Ignored if page, pageid or oldid is used'
 559                         ),
 560                         'uselang' => 'Which language to parse the request in',
 561                         'section' => 'Only retrieve the content of this section number',
 562                         'disablepp' => 'Disable the PP Report from the parser output',
 563                 );
 564         }
 565
 566         public function getDescription() {
 567                 return 'Parses wikitext and returns parser output';
 568         }
 569
 570         public function getPossibleErrors() {
 571                 return array_merge( parent::getPossibleErrors(), array(
 572                         array( 'code' => 'params', 'info' => 'The page parameter cannot be used together with the text and title parameters' ),
 573                         array( 'code' => 'missingrev', 'info' => 'There is no revision ID oldid' ),
 574                         array( 'code' => 'permissiondenied', 'info' => 'You don\'t have permission to view deleted revisions' ),
 575                         array( 'code' => 'missingtitle', 'info' => 'The page you specified doesn\'t exist' ),
 576                         array( 'code' => 'nosuchsection', 'info' => 'There is no section sectionnumber in page' ),
 577                         array( 'nosuchpageid' ),
 578                         array( 'invalidtitle', 'title' ),
 579                 ) );
 580         }
 581
 582         protected function getExamples() {
 583                 return array(
 584                         'api.php?action=parse&text={{Project:Sandbox}}'
 585                 );
 586         }
 587
 588         public function getVersion() {
 589                 return __CLASS__ . ': $Id$';
 590         }
 591 }