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