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                         $title = Title::newFromText( $link );
 328
 329                         $entry['lang'] = $bits[0];
 330                         if ( $title ) {
 331                                 $entry['url'] = $title->getFullURL();
 332                         }
 333                         $this->getResult()->setContent( $entry, $bits[1] );
 334                         $result[] = $entry;
 335                 }
 336                 return $result;
 337         }
 338
 339         private function formatCategoryLinks( $links ) {
 340                 $result = array();
 341                 foreach ( $links as $link => $sortkey ) {
 342                         $entry = array();
 343                         $entry['sortkey'] = $sortkey;
 344                         $this->getResult()->setContent( $entry, $link );
 345                         $result[] = $entry;
 346                 }
 347                 return $result;
 348         }
 349
 350         private function categoriesHtml( $categories ) {
 351                 global $wgOut, $wgUser;
 352                 $wgOut->addCategoryLinks( $categories );
 353                 $sk = $wgUser->getSkin();
 354                 return $sk->getCategories();
 355         }
 356
 357         private function languagesHtml( $languages ) {
 358                 global $wgOut, $wgUser;
 359                 $wgOut->setLanguageLinks( $languages );
 360                 $sk = $wgUser->getSkin();
 361                 return $sk->otherLanguages();
 362         }
 363
 364         private function formatLinks( $links ) {
 365                 $result = array();
 366                 foreach ( $links as $ns => $nslinks ) {
 367                         foreach ( $nslinks as $title => $id ) {
 368                                 $entry = array();
 369                                 $entry['ns'] = $ns;
 370                                 $this->getResult()->setContent( $entry, Title::makeTitle( $ns, $title )->getFullText() );
 371                                 if ( $id != 0 ) {
 372                                         $entry['exists'] = '';
 373                                 }
 374                                 $result[] = $entry;
 375                         }
 376                 }
 377                 return $result;
 378         }
 379
 380         private function formatIWLinks( $iw ) {
 381                 $result = array();
 382                 foreach ( $iw as $prefix => $titles ) {
 383                         foreach ( array_keys( $titles ) as $title ) {
 384                                 $entry = array();
 385                                 $entry['prefix'] = $prefix;
 386
 387                                 $title = Title::newFromText( "{$prefix}:{$title}" );
 388                                 if ( $title ) {
 389                                         $entry['url'] = $title->getFullURL();
 390                                 }
 391
 392                                 $this->getResult()->setContent( $entry, $title->getFullText() );
 393                                 $result[] = $entry;
 394                         }
 395                 }
 396                 return $result;
 397         }
 398
 399         private function formatHeadItems( $headItems ) {
 400                 $result = array();
 401                 foreach ( $headItems as $tag => $content ) {
 402                         $entry = array();
 403                         $entry['tag'] = $tag;
 404                         $this->getResult()->setContent( $entry, $content );
 405                         $result[] = $entry;
 406                 }
 407                 return $result;
 408         }
 409
 410         private function formatCss( $css ) {
 411                 $result = array();
 412                 foreach ( $css as $file => $link ) {
 413                         $entry = array();
 414                         $entry['file'] = $file;
 415                         $this->getResult()->setContent( $entry, $link );
 416                         $result[] = $entry;
 417                 }
 418                 return $result;
 419         }
 420
 421         private function setIndexedTagNames( &$array, $mapping ) {
 422                 foreach ( $mapping as $key => $name ) {
 423                         if ( isset( $array[$key] ) ) {
 424                                 $this->getResult()->setIndexedTagName( $array[$key], $name );
 425                         }
 426                 }
 427         }
 428
 429         public function getAllowedParams() {
 430                 return array(
 431                         'title' => array(
 432                                 ApiBase::PARAM_DFLT => 'API',
 433                         ),
 434                         'text' => null,
 435                         'summary' => null,
 436                         'page' => null,
 437                         'pageid' => null,
 438                         'redirects' => false,
 439                         'oldid' => null,
 440                         'prop' => array(
 441                                 ApiBase::PARAM_DFLT => 'text|langlinks|categories|links|templates|images|externallinks|sections|revid|displaytitle',
 442                                 ApiBase::PARAM_ISMULTI => true,
 443                                 ApiBase::PARAM_TYPE => array(
 444                                         'text',
 445                                         'langlinks',
 446                                         'languageshtml',
 447                                         'categories',
 448                                         'categorieshtml',
 449                                         'links',
 450                                         'templates',
 451                                         'images',
 452                                         'externallinks',
 453                                         'sections',
 454                                         'revid',
 455                                         'displaytitle',
 456                                         'headitems',
 457                                         'headhtml',
 458                                         'iwlinks',
 459                                 )
 460                         ),
 461                         'pst' => false,
 462                         'onlypst' => false,
 463                         'uselang' => null,
 464                         'section' => null,
 465                         'disablepp' => false,
 466                 );
 467         }
 468
 469         public function getParamDescription() {
 470                 $p = $this->getModulePrefix();
 471                 return array(
 472                         'text' => 'Wikitext to parse',
 473                         'summary' => 'Summary to parse',
 474                         'redirects' => "If the {$p}page parameter is set to a redirect, resolve it",
 475                         'title' => 'Title of page the text belongs to',
 476                         'page' => "Parse the content of this page. Cannot be used together with {$p}text and {$p}title",
 477                         'pageid' => "Parse the content of this page. Overrides {$p}page",
 478                         'oldid' => "Parse the content of this revision. Overrides {$p}page and {$p}pageid",
 479                         'prop' => array(
 480                                 'Which pieces of information to get',
 481                                 ' text           - Gives the parsed text of the wikitext',
 482                                 ' langlinks      - Gives the language links in the parsed wikitext',
 483                                 ' categories     - Gives the categories in the parsed wikitext',
 484                                 ' categorieshtml - Gives the HTML version of the categories',
 485                                 ' languageshtml  - Gives the HTML version of the language links',
 486                                 ' links          - Gives the internal links in the parsed wikitext',
 487                                 ' templates      - Gives the templates in the parsed wikitext',
 488                                 ' images         - Gives the images in the parsed wikitext',
 489                                 ' externallinks  - Gives the external links in the parsed wikitext',
 490                                 ' sections       - Gives the sections in the parsed wikitext',
 491                                 ' revid          - Adds the revision ID of the parsed page',
 492                                 ' displaytitle   - Adds the title of the parsed wikitext',
 493                                 ' headitems      - Gives items to put in the <head> of the page',
 494                                 ' headhtml       - Gives parsed <head> of the page',
 495                                 ' iwlinks        - Gives interwiki links in the parsed wikitext',
 496                                 'NOTE: Section tree is only generated if there are more than 4 sections, or if the __TOC__ keyword is present'
 497                         ),
 498                         'pst' => array(
 499                                 'Do a pre-save transform on the input before parsing it',
 500                                 'Ignored if page, pageid or oldid is used'
 501                         ),
 502                         'onlypst' => array(
 503                                 'Do a pre-save transform (PST) on the input, but don\'t parse it',
 504                                 'Returns the same wikitext, after a PST has been applied. Ignored if page, pageid or oldid is used'
 505                         ),
 506                         'uselang' => 'Which language to parse the request in',
 507                         'section' => 'Only retrieve the content of this section number',
 508                         'disablepp' => 'Disable the PP Report from the parser output',
 509                 );
 510         }
 511
 512         public function getDescription() {
 513                 return 'This module parses wikitext and returns parser output';
 514         }
 515
 516         public function getPossibleErrors() {
 517                 return array_merge( parent::getPossibleErrors(), array(
 518                         array( 'code' => 'params', 'info' => 'The page parameter cannot be used together with the text and title parameters' ),
 519                         array( 'code' => 'missingrev', 'info' => 'There is no revision ID oldid' ),
 520                         array( 'code' => 'permissiondenied', 'info' => 'You don\'t have permission to view deleted revisions' ),
 521                         array( 'code' => 'missingtitle', 'info' => 'The page you specified doesn\'t exist' ),
 522                         array( 'code' => 'nosuchsection', 'info' => 'There is no section sectionnumber in page' ),
 523                         array( 'nosuchpageid' ),
 524                 ) );
 525         }
 526
 527         protected function getExamples() {
 528                 return array(
 529                         'api.php?action=parse&text={{Project:Sandbox}}'
 530                 );
 531         }
 532
 533         public function getVersion() {
 534                 return __CLASS__ . ': $Id$';
 535         }
 536 }