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, $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
  81                 $redirValues = null;
  82
  83                 if ( !is_null( $oldid ) || !is_null( $pageid ) || !is_null( $page ) ) {
  84
  85                         if ( !is_null( $oldid ) ) {
  86                                 // Don't use the parser cache
  87                                 $rev = Revision::newFromID( $oldid );
  88                                 if ( !$rev ) {
  89                                         $this->dieUsage( "There is no revision ID $oldid", 'missingrev' );
  90                                 }
  91                                 if ( !$rev->userCan( Revision::DELETED_TEXT ) ) {
  92                                         $this->dieUsage( "You don't have permission to view deleted revisions", 'permissiondenied' );
  93                                 }
  94
  95                                 $titleObj = $rev->getTitle();
  96
  97                                 $wgTitle = $titleObj;
  98
  99                                 // If for some reason the "oldid" is actually the current revision, it may be cached
 100                                 if ( $titleObj->getLatestRevID() === intval( $oldid ) )  {
 101                                         $articleObj = new Article( $titleObj, 0 );
 102
 103                                         $p_result = $this->getParsedSectionOrText( $articleObj, $titleObj, $popts, $pageid ) ;
 104
 105                                 } else { // This is an old revision, so get the text differently
 106                                         $text = $rev->getText( Revision::FOR_THIS_USER );
 107
 108                                         $wgTitle = $titleObj;
 109
 110                                         if ( $this->section !== false ) {
 111                                                 $text = $this->getSectionText( $text, 'r' . $rev->getId() );
 112                                         }
 113
 114                                         $p_result = $wgParser->parse( $text, $titleObj, $popts );
 115                                 }
 116
 117                         } else { // Not $oldid
 118
 119                                 if ( !is_null ( $pageid ) ) {
 120                                         $titleObj = Title::newFromID( $pageid );
 121
 122                                         if ( !$titleObj ) {
 123                                                 $this->dieUsageMsg( array( 'nosuchpageid', $pageid ) );
 124                                         }
 125                                 } else { // $page
 126
 127                                         if ( $params['redirects'] ) {
 128                                                 $req = new FauxRequest( array(
 129                                                         'action' => 'query',
 130                                                         'redirects' => '',
 131                                                         'titles' => $page
 132                                                 ) );
 133                                                 $main = new ApiMain( $req );
 134                                                 $main->execute();
 135                                                 $data = $main->getResultData();
 136                                                 $redirValues = @$data['query']['redirects'];
 137                                                 $to = $page;
 138                                                 foreach ( (array)$redirValues as $r ) {
 139                                                         $to = $r['to'];
 140                                                 }
 141                                         } else {
 142                                                 $to = $page;
 143                                         }
 144                                         $titleObj = Title::newFromText( $to );
 145                                         if ( !$titleObj || !$titleObj->exists() ) {
 146                                                 $this->dieUsage( "The page you specified doesn't exist", 'missingtitle' );
 147                                         }
 148                                 }
 149                                 $wgTitle = $titleObj;
 150
 151                                 $articleObj = new Article( $titleObj, 0 );
 152                                 if ( isset( $prop['revid'] ) ) {
 153                                         $oldid = $articleObj->getRevIdFetched();
 154                                 }
 155
 156                                 $p_result = $this->getParsedSectionOrText( $articleObj, $titleObj, $popts, $pageid ) ;
 157                         }
 158
 159                 } else { // Not $oldid, $pageid, $page. Hence based on $text
 160
 161                         $titleObj = Title::newFromText( $title );
 162                         if ( !$titleObj ) {
 163                                 $titleObj = Title::newFromText( 'API' );
 164                         }
 165                         $wgTitle = $titleObj;
 166
 167                         if ( $this->section !== false ) {
 168                                 $text = $this->getSectionText( $text, $titleObj->getText() );
 169                         }
 170
 171                         if ( $params['pst'] || $params['onlypst'] ) {
 172                                 $text = $wgParser->preSaveTransform( $text, $titleObj, $wgUser, $popts );
 173                         }
 174                         if ( $params['onlypst'] ) {
 175                                 // Build a result and bail out
 176                                 $result_array['text'] = array();
 177                                 $this->getResult()->setContent( $result_array['text'], $text );
 178                                 $this->getResult()->addValue( null, $this->getModuleName(), $result_array );
 179                                 return;
 180                         }
 181                         $p_result = $wgParser->parse( $text, $titleObj, $popts );
 182                 }
 183
 184                 // Return result
 185                 $result = $this->getResult();
 186                 $result_array = array();
 187
 188                 $result_array['title'] = $titleObj->getPrefixedText();
 189
 190                 if ( !is_null( $oldid ) ) {
 191                         $result_array['revid'] = intval( $oldid );
 192                 }
 193
 194                 if ( $params['redirects'] && !is_null( $redirValues ) ) {
 195                         $result_array['redirects'] = $redirValues;
 196                 }
 197
 198                 if ( isset( $prop['text'] ) ) {
 199                         $result_array['text'] = array();
 200                         $result->setContent( $result_array['text'], $p_result->getText() );
 201                 }
 202
 203                 if ( !is_null( $params['summary'] ) ) {
 204                         $result_array['parsedsummary'] = array();
 205                         $result->setContent( $result_array['parsedsummary'], $wgUser->getSkin()->formatComment( $params['summary'], $titleObj ) );
 206                 }
 207
 208                 if ( isset( $prop['langlinks'] ) ) {
 209                         $result_array['langlinks'] = $this->formatLangLinks( $p_result->getLanguageLinks() );
 210                 }
 211                 if ( isset( $prop['languageshtml'] ) ) {
 212                         $languagesHtml = $this->languagesHtml( $p_result->getLanguageLinks() );
 213                         $result_array['languageshtml'] = array();
 214                         $result->setContent( $result_array['languageshtml'], $languagesHtml );
 215                 }
 216                 if ( isset( $prop['categories'] ) ) {
 217                         $result_array['categories'] = $this->formatCategoryLinks( $p_result->getCategories() );
 218                 }
 219                 if ( isset( $prop['categorieshtml'] ) ) {
 220                         $categoriesHtml = $this->categoriesHtml( $p_result->getCategories() );
 221                         $result_array['categorieshtml'] = array();
 222                         $result->setContent( $result_array['categorieshtml'], $categoriesHtml );
 223                 }
 224                 if ( isset( $prop['links'] ) ) {
 225                         $result_array['links'] = $this->formatLinks( $p_result->getLinks() );
 226                 }
 227                 if ( isset( $prop['templates'] ) ) {
 228                         $result_array['templates'] = $this->formatLinks( $p_result->getTemplates() );
 229                 }
 230                 if ( isset( $prop['images'] ) ) {
 231                         $result_array['images'] = array_keys( $p_result->getImages() );
 232                 }
 233                 if ( isset( $prop['externallinks'] ) ) {
 234                         $result_array['externallinks'] = array_keys( $p_result->getExternalLinks() );
 235                 }
 236                 if ( isset( $prop['sections'] ) ) {
 237                         $result_array['sections'] = $p_result->getSections();
 238                 }
 239
 240                 if ( isset( $prop['displaytitle'] ) ) {
 241                         $result_array['displaytitle'] = $p_result->getDisplayTitle() ?
 242                                                         $p_result->getDisplayTitle() :
 243                                                         $titleObj->getPrefixedText();
 244                 }
 245
 246                 if ( isset( $prop['headitems'] ) || isset( $prop['headhtml'] ) ) {
 247                         $out = new OutputPage;
 248                         $out->addParserOutputNoText( $p_result );
 249                         $userSkin = $wgUser->getSkin();
 250                 }
 251
 252                 if ( isset( $prop['headitems'] ) ) {
 253                         $headItems = $this->formatHeadItems( $p_result->getHeadItems() );
 254
 255                         $userSkin->setupUserCss( $out );
 256                         $css = $this->formatCss( $out->buildCssLinksArray() );
 257
 258                         $scripts = array( $out->getHeadScripts( $userSkin ) );
 259
 260                         $result_array['headitems'] = array_merge( $headItems, $css, $scripts );
 261                 }
 262
 263                 if ( isset( $prop['headhtml'] ) ) {
 264                         $result_array['headhtml'] = array();
 265                         $result->setContent( $result_array['headhtml'], $out->headElement( $userSkin ) );
 266                 }
 267
 268                 if ( isset( $prop['iwlinks'] ) ) {
 269                         $result_array['iwlinks'] = $this->formatIWLinks( $p_result->getInterwikiLinks() );
 270                 }
 271
 272                 $result_mapping = array(
 273                         'redirects' => 'r',
 274                         'langlinks' => 'll',
 275                         'categories' => 'cl',
 276                         'links' => 'pl',
 277                         'templates' => 'tl',
 278                         'images' => 'img',
 279                         'externallinks' => 'el',
 280                         'iwlinks' => 'iw',
 281                         'sections' => 's',
 282                         'headitems' => 'hi',
 283                 );
 284                 $this->setIndexedTagNames( $result_array, $result_mapping );
 285                 $result->addValue( null, $this->getModuleName(), $result_array );
 286
 287                 if ( !is_null( $oldLang ) ) {
 288                         $wgLang = $oldLang; // Reset $wgLang to $oldLang
 289                 }
 290         }
 291
 292         /**
 293          * @param  $articleObj Article
 294          * @param  $titleObj Title
 295          * @param  $popts ParserOptions
 296          * @param  $pageId Int
 297          * @return ParserOutput
 298          */
 299         private function getParsedSectionOrText( $articleObj, $titleObj, $popts, $pageId = null ) {
 300                 if ( $this->section !== false ) {
 301                         global $wgParser;
 302
 303                         $text = $this->getSectionText( $articleObj->getRawText(), !is_null ( $pageId )
 304                                         ? 'page id ' . $pageId : $titleObj->getText() );
 305
 306                         return $wgParser->parse( $text, $titleObj, $popts );
 307                 } else {
 308                         // Try the parser cache first
 309                         return $articleObj->getParserOutput();
 310                 }
 311         }
 312
 313         private function getSectionText( $text, $what ) {
 314                 global $wgParser;
 315                 $text = $wgParser->getSection( $text, $this->section, false );
 316                 if ( $text === false ) {
 317                         $this->dieUsage( "There is no section {$this->section} in " . $what, 'nosuchsection' );
 318                 }
 319                 return $text;
 320         }
 321
 322         private function formatLangLinks( $links ) {
 323                 $result = array();
 324                 foreach ( $links as $link ) {
 325                         $entry = array();
 326                         $bits = explode( ':', $link, 2 );
 327                         $entry['lang'] = $bits[0];
 328                         $this->getResult()->setContent( $entry, $bits[1] );
 329                         $result[] = $entry;
 330                 }
 331                 return $result;
 332         }
 333
 334         private function formatCategoryLinks( $links ) {
 335                 $result = array();
 336                 foreach ( $links as $link => $sortkey ) {
 337                         $entry = array();
 338                         $entry['sortkey'] = $sortkey;
 339                         $this->getResult()->setContent( $entry, $link );
 340                         $result[] = $entry;
 341                 }
 342                 return $result;
 343         }
 344
 345         private function categoriesHtml( $categories ) {
 346                 global $wgOut, $wgUser;
 347                 $wgOut->addCategoryLinks( $categories );
 348                 $sk = $wgUser->getSkin();
 349                 return $sk->getCategories();
 350         }
 351
 352         private function languagesHtml( $languages ) {
 353                 global $wgOut, $wgUser;
 354                 $wgOut->setLanguageLinks( $languages );
 355                 $sk = $wgUser->getSkin();
 356                 return $sk->otherLanguages();
 357         }
 358
 359         private function formatLinks( $links ) {
 360                 $result = array();
 361                 foreach ( $links as $ns => $nslinks ) {
 362                         foreach ( $nslinks as $title => $id ) {
 363                                 $entry = array();
 364                                 $entry['ns'] = $ns;
 365                                 $this->getResult()->setContent( $entry, Title::makeTitle( $ns, $title )->getFullText() );
 366                                 if ( $id != 0 ) {
 367                                         $entry['exists'] = '';
 368                                 }
 369                                 $result[] = $entry;
 370                         }
 371                 }
 372                 return $result;
 373         }
 374
 375         private function formatIWLinks( $iw ) {
 376                 $result = array();
 377                 foreach ( $iw as $prefix => $titles ) {
 378                         foreach ( array_keys( $titles ) as $title ) {
 379                                 $entry = array();
 380                                 $entry['prefix'] = $prefix;
 381
 382                                 $title = Title::newFromText( "{$prefix}:{$title}" );
 383                                 if ( $title ) {
 384                                         $entry['url'] = $title->getFullURL();
 385                                 }
 386
 387                                 $this->getResult()->setContent( $entry, $title->getFullText() );
 388                                 $result[] = $entry;
 389                         }
 390                 }
 391                 return $result;
 392         }
 393
 394         private function formatHeadItems( $headItems ) {
 395                 $result = array();
 396                 foreach ( $headItems as $tag => $content ) {
 397                         $entry = array();
 398                         $entry['tag'] = $tag;
 399                         $this->getResult()->setContent( $entry, $content );
 400                         $result[] = $entry;
 401                 }
 402                 return $result;
 403         }
 404
 405         private function formatCss( $css ) {
 406                 $result = array();
 407                 foreach ( $css as $file => $link ) {
 408                         $entry = array();
 409                         $entry['file'] = $file;
 410                         $this->getResult()->setContent( $entry, $link );
 411                         $result[] = $entry;
 412                 }
 413                 return $result;
 414         }
 415
 416         private function setIndexedTagNames( &$array, $mapping ) {
 417                 foreach ( $mapping as $key => $name ) {
 418                         if ( isset( $array[$key] ) ) {
 419                                 $this->getResult()->setIndexedTagName( $array[$key], $name );
 420                         }
 421                 }
 422         }
 423
 424         public function getAllowedParams() {
 425                 return array(
 426                         'title' => array(
 427                                 ApiBase::PARAM_DFLT => 'API',
 428                         ),
 429                         'text' => null,
 430                         'summary' => null,
 431                         'page' => null,
 432                         'pageid' => null,
 433                         'redirects' => false,
 434                         'oldid' => null,
 435                         'prop' => array(
 436                                 ApiBase::PARAM_DFLT => 'text|langlinks|categories|links|templates|images|externallinks|sections|revid|displaytitle',
 437                                 ApiBase::PARAM_ISMULTI => true,
 438                                 ApiBase::PARAM_TYPE => array(
 439                                         'text',
 440                                         'langlinks',
 441                                         'languageshtml',
 442                                         'categories',
 443                                         'categorieshtml',
 444                                         'links',
 445                                         'templates',
 446                                         'images',
 447                                         'externallinks',
 448                                         'sections',
 449                                         'revid',
 450                                         'displaytitle',
 451                                         'headitems',
 452                                         'headhtml',
 453                                         'iwlinks',
 454                                 )
 455                         ),
 456                         'pst' => false,
 457                         'onlypst' => false,
 458                         'uselang' => null,
 459                         'section' => null,
 460                         'disablepp' => false,
 461                 );
 462         }
 463
 464         public function getParamDescription() {
 465                 $p = $this->getModulePrefix();
 466                 return array(
 467                         'text' => 'Wikitext to parse',
 468                         'summary' => 'Summary to parse',
 469                         'redirects' => "If the {$p}page parameter is set to a redirect, resolve it",
 470                         'title' => 'Title of page the text belongs to',
 471                         'page' => "Parse the content of this page. Cannot be used together with {$p}text and {$p}title",
 472                         'pageid' => "Parse the content of this page. Overrides {$p}page",
 473                         'oldid' => "Parse the content of this revision. Overrides {$p}page and {$p}pageid",
 474                         'prop' => array(
 475                                 'Which pieces of information to get',
 476                                 ' text           - Gives the parsed text of the wikitext',
 477                                 ' langlinks      - Gives the language links in the parsed wikitext',
 478                                 ' categories     - Gives the categories in the parsed wikitext',
 479                                 ' categorieshtml - Gives the HTML version of the categories',
 480                                 ' languageshtml  - Gives the HTML version of the language links',
 481                                 ' links          - Gives the internal links in the parsed wikitext',
 482                                 ' templates      - Gives the templates in the parsed wikitext',
 483                                 ' images         - Gives the images in the parsed wikitext',
 484                                 ' externallinks  - Gives the external links in the parsed wikitext',
 485                                 ' sections       - Gives the sections in the parsed wikitext',
 486                                 ' revid          - Adds the revision ID of the parsed page',
 487                                 ' displaytitle   - Adds the title of the parsed wikitext',
 488                                 ' headitems      - Gives items to put in the <head> of the page',
 489                                 ' headhtml       - Gives parsed <head> of the page',
 490                                 ' iwlinks        - Gives interwiki links in the parsed wikitext',
 491                                 'NOTE: Section tree is only generated if there are more than 4 sections, or if the __TOC__ keyword is present'
 492                         ),
 493                         'pst' => array(
 494                                 'Do a pre-save transform on the input before parsing it',
 495                                 'Ignored if page, pageid or oldid is used'
 496                         ),
 497                         'onlypst' => array(
 498                                 'Do a pre-save transform (PST) on the input, but don\'t parse it',
 499                                 'Returns the same wikitext, after a PST has been applied. Ignored if page, pageid or oldid is used'
 500                         ),
 501                         'uselang' => 'Which language to parse the request in',
 502                         'section' => 'Only retrieve the content of this section number',
 503                         'disablepp' => 'Disable the PP Report from the parser output',
 504                 );
 505         }
 506
 507         public function getDescription() {
 508                 return 'This module parses wikitext and returns parser output';
 509         }
 510
 511         public function getPossibleErrors() {
 512                 return array_merge( parent::getPossibleErrors(), array(
 513                         array( 'code' => 'params', 'info' => 'The page parameter cannot be used together with the text and title parameters' ),
 514                         array( 'code' => 'missingrev', 'info' => 'There is no revision ID oldid' ),
 515                         array( 'code' => 'permissiondenied', 'info' => 'You don\'t have permission to view deleted revisions' ),
 516                         array( 'code' => 'missingtitle', 'info' => 'The page you specified doesn\'t exist' ),
 517                         array( 'code' => 'nosuchsection', 'info' => 'There is no section sectionnumber in page' ),
 518                         array( 'nosuchpageid' ),
 519                 ) );
 520         }
 521
 522         protected function getExamples() {
 523                 return array(
 524                         'api.php?action=parse&text={{Project:Sandbox}}'
 525                 );
 526         }
 527
 528         public function getVersion() {
 529                 return __CLASS__ . ': $Id$';
 530         }
 531 }