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