includes/api/ApiParse.php

   1 <?php
   2 /**
   3  * API for MediaWiki 1.8+
   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() === $oldid ) {
  98                                         $p_result = false;
  99                                         $pcache = ParserCache::singleton();
 100                                         if ( $wgEnableParserCache ) {
 101                                                 $p_result = $pcache->get( $titleObj, $popts );
 102                                         }
 103                                         if ( !$p_result ) {
 104                                                 $text = $rev->getText( Revision::FOR_THIS_USER );
 105                                                 $p_result = $wgParser->parse( $text, $titleObj, $popts );
 106
 107                                                 if ( $wgEnableParserCache ) {
 108                                                         $pcache->save( $p_result, $titleObj, $popts );
 109                                                 }
 110                                         }
 111                                 } else {
 112                                         $text = $rev->getText( Revision::FOR_THIS_USER );
 113
 114                                         $wgTitle = $titleObj;
 115
 116                                         if ( $this->section !== false ) {
 117                                                 $text = $this->getSectionText( $text, 'r' . $rev );
 118                                         }
 119
 120                                         $p_result = $wgParser->parse( $text, $titleObj, $popts );
 121                                 }
 122                         } else {
 123                                 if ( !is_null ( $pageid ) ) {
 124                                         $titleObj = Title::newFromID( $pageid );
 125
 126                                         if ( !$titleObj ) {
 127                                                 $this->dieUsageMsg( array( 'nosuchpageid', $pageid ) );
 128                                         }
 129                                 } else {
 130                                         if ( $params['redirects'] ) {
 131                                                 $req = new FauxRequest( array(
 132                                                         'action' => 'query',
 133                                                         'redirects' => '',
 134                                                         'titles' => $page
 135                                                 ) );
 136                                                 $main = new ApiMain( $req );
 137                                                 $main->execute();
 138                                                 $data = $main->getResultData();
 139                                                 $redirValues = @$data['query']['redirects'];
 140                                                 $to = $page;
 141                                                 foreach ( (array)$redirValues as $r ) {
 142                                                         $to = $r['to'];
 143                                                 }
 144                                         } else {
 145                                                 $to = $page;
 146                                         }
 147                                         $titleObj = Title::newFromText( $to );
 148                                         if ( !$titleObj ) {
 149                                                 $this->dieUsage( "The page you specified doesn't exist", 'missingtitle' );
 150                                         }
 151                                 }
 152                                 $wgTitle = $titleObj;
 153
 154                                 $articleObj = new Article( $titleObj );
 155                                 if ( isset( $prop['revid'] ) ) {
 156                                         $oldid = $articleObj->getRevIdFetched();
 157                                 }
 158
 159                                 if ( $this->section !== false ) {
 160                                         $text = $this->getSectionText( $text, !is_null ( $pageid ) ? 'page id ' . $pageid : $titleObj->getText() );
 161                                         $p_result = $wgParser->parse( $text, $titleObj, $popts );
 162                                 } else {
 163                                         // Try the parser cache first
 164                                         $p_result = false;
 165                                         $pcache = ParserCache::singleton();
 166                                         if ( $wgEnableParserCache ) {
 167                                                 $p_result = $pcache->get( $articleObj, $popts );
 168                                         }
 169                                         if ( !$p_result ) {
 170                                                 $p_result = $wgParser->parse( $articleObj->getContent(), $titleObj, $popts );
 171
 172                                                 if ( $wgEnableParserCache ) {
 173                                                         $pcache->save( $p_result, $articleObj, $popts );
 174                                                 }
 175                                         }
 176                                 }
 177                         }
 178                 } else {
 179                         $titleObj = Title::newFromText( $title );
 180                         if ( !$titleObj ) {
 181                                 $titleObj = Title::newFromText( 'API' );
 182                         }
 183                         $wgTitle = $titleObj;
 184
 185                         if ( $this->section !== false ) {
 186                                 $text = $this->getSectionText( $text, $titleObj->getText() );
 187                         }
 188
 189                         if ( $params['pst'] || $params['onlypst'] ) {
 190                                 $text = $wgParser->preSaveTransform( $text, $titleObj, $wgUser, $popts );
 191                         }
 192                         if ( $params['onlypst'] ) {
 193                                 // Build a result and bail out
 194                                 $result_array['text'] = array();
 195                                 $this->getResult()->setContent( $result_array['text'], $text );
 196                                 $this->getResult()->addValue( null, $this->getModuleName(), $result_array );
 197                                 return;
 198                         }
 199                         $p_result = $wgParser->parse( $text, $titleObj, $popts );
 200                 }
 201
 202                 // Return result
 203                 $result = $this->getResult();
 204                 $result_array = array();
 205                 if ( $params['redirects'] && !is_null( $redirValues ) ) {
 206                         $result_array['redirects'] = $redirValues;
 207                 }
 208
 209                 if ( isset( $prop['text'] ) ) {
 210                         $result_array['text'] = array();
 211                         $result->setContent( $result_array['text'], $p_result->getText() );
 212                 }
 213
 214                 if ( !is_null( $params['summary'] ) ) {
 215                         $result_array['parsedsummary'] = array();
 216                         $result->setContent( $result_array['parsedsummary'], $wgUser->getSkin()->formatComment( $params['summary'], $titleObj ) );
 217                 }
 218
 219                 if ( isset( $prop['langlinks'] ) ) {
 220                         $result_array['langlinks'] = $this->formatLangLinks( $p_result->getLanguageLinks() );
 221                 }
 222                 if ( isset( $prop['categories'] ) ) {
 223                         $result_array['categories'] = $this->formatCategoryLinks( $p_result->getCategories() );
 224                 }
 225                 if ( isset( $prop['links'] ) ) {
 226                         $result_array['links'] = $this->formatLinks( $p_result->getLinks() );
 227                 }
 228                 if ( isset( $prop['templates'] ) ) {
 229                         $result_array['templates'] = $this->formatLinks( $p_result->getTemplates() );
 230                 }
 231                 if ( isset( $prop['images'] ) ) {
 232                         $result_array['images'] = array_keys( $p_result->getImages() );
 233                 }
 234                 if ( isset( $prop['externallinks'] ) ) {
 235                         $result_array['externallinks'] = array_keys( $p_result->getExternalLinks() );
 236                 }
 237                 if ( isset( $prop['sections'] ) ) {
 238                         $result_array['sections'] = $p_result->getSections();
 239                 }
 240
 241                 if ( isset( $prop['displaytitle'] ) ) {
 242                         $result_array['displaytitle'] = $p_result->getDisplayTitle() ?
 243                                                         $p_result->getDisplayTitle() :
 244                                                         $titleObj->getPrefixedText();
 245                 }
 246
 247                 if ( isset( $prop['headitems'] ) || isset( $prop['headhtml'] ) ) {
 248                         $out = new OutputPage;
 249                         $out->addParserOutputNoText( $p_result );
 250                         $userSkin = $wgUser->getSkin();
 251                 }
 252
 253                 if ( isset( $prop['headitems'] ) ) {
 254                         $headItems = $this->formatHeadItems( $p_result->getHeadItems() );
 255
 256                         $userSkin->setupUserCss( $out );
 257                         $css = $this->formatCss( $out->buildCssLinksArray() );
 258
 259                         $scripts = array( $out->getHeadScripts( $userSkin ) );
 260
 261                         $result_array['headitems'] = array_merge( $headItems, $css, $scripts );
 262                 }
 263
 264                 if ( isset( $prop['headhtml'] ) ) {
 265                         $result_array['headhtml'] = array();
 266                         $result->setContent( $result_array['headhtml'], $out->headElement( $userSkin ) );
 267                 }
 268
 269                 if ( isset( $prop['iwlinks'] ) ) {
 270                         $result_array['iwlinks'] = $this->formatIWLinks( $p_result->getInterwikiLinks() );
 271                 }
 272
 273                 if ( !is_null( $oldid ) ) {
 274                         $result_array['revid'] = intval( $oldid );
 275                 }
 276
 277                 $result_mapping = array(
 278                         'redirects' => 'r',
 279                         'langlinks' => 'll',
 280                         'categories' => 'cl',
 281                         'links' => 'pl',
 282                         'templates' => 'tl',
 283                         'images' => 'img',
 284                         'externallinks' => 'el',
 285                         'iwlinks' => 'iw',
 286                         'sections' => 's',
 287                         'headitems' => 'hi',
 288                 );
 289                 $this->setIndexedTagNames( $result_array, $result_mapping );
 290                 $result->addValue( null, $this->getModuleName(), $result_array );
 291
 292                 if ( !is_null( $oldLang ) ) {
 293                         $wgLang = $oldLang; // Reset $wgLang to $oldLang
 294                 }
 295         }
 296
 297         private function getSectionText( $text, $what ) {
 298                 global $wgParser;
 299                 $text = $wgParser->getSection( $text, $this->section, false );
 300                 if ( $text === false ) {
 301                         $this->dieUsage( "There is no section {$this->section} in " . $what, 'nosuchsection' );
 302                 }
 303                 return $text;
 304         }
 305
 306         private function formatLangLinks( $links ) {
 307                 $result = array();
 308                 foreach ( $links as $link ) {
 309                         $entry = array();
 310                         $bits = explode( ':', $link, 2 );
 311                         $entry['lang'] = $bits[0];
 312                         $this->getResult()->setContent( $entry, $bits[1] );
 313                         $result[] = $entry;
 314                 }
 315                 return $result;
 316         }
 317
 318         private function formatCategoryLinks( $links ) {
 319                 $result = array();
 320                 foreach ( $links as $link => $sortkey ) {
 321                         $entry = array();
 322                         $entry['sortkey'] = $sortkey;
 323                         $this->getResult()->setContent( $entry, $link );
 324                         $result[] = $entry;
 325                 }
 326                 return $result;
 327         }
 328
 329         private function formatLinks( $links ) {
 330                 $result = array();
 331                 foreach ( $links as $ns => $nslinks ) {
 332                         foreach ( $nslinks as $title => $id ) {
 333                                 $entry = array();
 334                                 $entry['ns'] = $ns;
 335                                 $this->getResult()->setContent( $entry, Title::makeTitle( $ns, $title )->getFullText() );
 336                                 if ( $id != 0 ) {
 337                                         $entry['exists'] = '';
 338                                 }
 339                                 $result[] = $entry;
 340                         }
 341                 }
 342                 return $result;
 343         }
 344
 345         private function formatIWLinks( $iw ) {
 346                 $result = array();
 347                 foreach ( $iw as $prefix => $titles ) {
 348                         foreach ( array_keys( $titles ) as $title ) {
 349                                 $entry = array();
 350                                 $entry['prefix'] = $prefix;
 351
 352                                 $title = Title::newFromText( "{$prefix}:{$title}" );
 353                                 if ( $title ) {
 354                                         $entry['url'] = $title->getFullURL();
 355                                 }
 356
 357                                 $this->getResult()->setContent( $entry, $title->getFullText() );
 358                                 $result[] = $entry;
 359                         }
 360                 }
 361                 return $result;
 362         }
 363
 364         private function formatHeadItems( $headItems ) {
 365                 $result = array();
 366                 foreach ( $headItems as $tag => $content ) {
 367                         $entry = array();
 368                         $entry['tag'] = $tag;
 369                         $this->getResult()->setContent( $entry, $content );
 370                         $result[] = $entry;
 371                 }
 372                 return $result;
 373         }
 374
 375         private function formatCss( $css ) {
 376                 $result = array();
 377                 foreach ( $css as $file => $link ) {
 378                         $entry = array();
 379                         $entry['file'] = $file;
 380                         $this->getResult()->setContent( $entry, $link );
 381                         $result[] = $entry;
 382                 }
 383                 return $result;
 384         }
 385
 386         private function setIndexedTagNames( &$array, $mapping ) {
 387                 foreach ( $mapping as $key => $name ) {
 388                         if ( isset( $array[$key] ) ) {
 389                                 $this->getResult()->setIndexedTagName( $array[$key], $name );
 390                         }
 391                 }
 392         }
 393
 394         public function getAllowedParams() {
 395                 return array(
 396                         'title' => array(
 397                                 ApiBase::PARAM_DFLT => 'API',
 398                         ),
 399                         'text' => null,
 400                         'summary' => null,
 401                         'page' => null,
 402                         'pageid' => null,
 403                         'redirects' => false,
 404                         'oldid' => null,
 405                         'prop' => array(
 406                                 ApiBase::PARAM_DFLT => 'text|langlinks|categories|links|templates|images|externallinks|sections|revid|displaytitle',
 407                                 ApiBase::PARAM_ISMULTI => true,
 408                                 ApiBase::PARAM_TYPE => array(
 409                                         'text',
 410                                         'langlinks',
 411                                         'categories',
 412                                         'links',
 413                                         'templates',
 414                                         'images',
 415                                         'externallinks',
 416                                         'sections',
 417                                         'revid',
 418                                         'displaytitle',
 419                                         'headitems',
 420                                         'headhtml',
 421                                         'iwlinks',
 422                                 )
 423                         ),
 424                         'pst' => false,
 425                         'onlypst' => false,
 426                         'uselang' => null,
 427                         'section' => null,
 428                         'disablepp' => false,
 429                 );
 430         }
 431
 432         public function getParamDescription() {
 433                 $p = $this->getModulePrefix();
 434                 return array(
 435                         'text' => 'Wikitext to parse',
 436                         'summary' => 'Summary to parse',
 437                         'redirects' => "If the {$p}page parameter is set to a redirect, resolve it",
 438                         'title' => 'Title of page the text belongs to',
 439                         'page' => "Parse the content of this page. Cannot be used together with {$p}text and {$p}title",
 440                         'pageid' => "Parse the content of this page. Overrides {$p}page",
 441                         'oldid' => "Parse the content of this revision. Overrides {$p}page and {$p}pageid",
 442                         'prop' => array(
 443                                 'Which pieces of information to get',
 444                                 ' text           - Gives the parsed text of the wikitext',
 445                                 ' langlinks      - Gives the langlinks the parsed wikitext',
 446                                 ' categories     - Gives the categories of the parsed wikitext',
 447                                 ' links          - Gives the internal links in the parsed wikitext',
 448                                 ' templates      - Gives the templates in the parsed wikitext',
 449                                 ' images         - Gives the images in the parsed wikitext',
 450                                 ' externallinks  - Gives the external links in the parsed wikitext',
 451                                 ' sections       - Gives the sections in the parsed wikitext',
 452                                 ' revid          - Adds the revision id of the parsed page',
 453                                 ' displaytitle   - Adds the title of the parsed wikitext',
 454                                 ' headitems      - Gives items to put in the <head> of the page',
 455                                 ' headhtml       - Gives parsed <head> of the page',
 456                                 ' iwlinks        - Gives interwiki links in the parsed wikitext',
 457                                 'NOTE: Section tree is only generated if there are more than 4 sections, or if the __TOC__ keyword is present'
 458                         ),
 459                         'pst' => array(
 460                                 'Do a pre-save transform on the input before parsing it',
 461                                 'Ignored if page, pageid or oldid is used'
 462                         ),
 463                         'onlypst' => array(
 464                                 'Do a pre-save transform (PST) on the input, but don\'t parse it',
 465                                 'Returns the same wikitext, after a PST has been applied. Ignored if page, pageid or oldid is used'
 466                         ),
 467                         'uselang' => 'Which language to parse the request in',
 468                         'section' => 'Only retrieve the content of this section number',
 469                         'disablepp' => 'Disable the PP Report from the parser output',
 470                 );
 471         }
 472
 473         public function getDescription() {
 474                 return 'This module parses wikitext and returns parser output';
 475         }
 476
 477         public function getPossibleErrors() {
 478                 return array_merge( parent::getPossibleErrors(), array(
 479                         array( 'code' => 'params', 'info' => 'The page parameter cannot be used together with the text and title parameters' ),
 480                         array( 'code' => 'missingrev', 'info' => 'There is no revision ID oldid' ),
 481                         array( 'code' => 'permissiondenied', 'info' => 'You don\'t have permission to view deleted revisions' ),
 482                         array( 'code' => 'missingtitle', 'info' => 'The page you specified doesn\'t exist' ),
 483                         array( 'code' => 'nosuchsection', 'info' => 'There is no section sectionnumber in page' ),
 484                         array( 'nosuchpageid' ),
 485                 ) );
 486         }
 487
 488         protected function getExamples() {
 489                 return array(
 490                         'api.php?action=parse&text={{Project:Sandbox}}'
 491                 );
 492         }
 493
 494         public function getVersion() {
 495                 return __CLASS__ . ': $Id$';
 496         }
 497 }