includes/api/ApiParse.php

   1 <?php
   2 /**
   3  * Created on Dec 01, 2007
   4  *
   5  * Copyright © 2007 Yuri Astrakhan <Firstname><Lastname>@gmail.com
   6  *
   7  * This program is free software; you can redistribute it and/or modify
   8  * it under the terms of the GNU General Public License as published by
   9  * the Free Software Foundation; either version 2 of the License, or
  10  * (at your option) any later version.
  11  *
  12  * This program is distributed in the hope that it will be useful,
  13  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  14  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  15  * GNU General Public License for more details.
  16  *
  17  * You should have received a copy of the GNU General Public License along
  18  * with this program; if not, write to the Free Software Foundation, Inc.,
  19  * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
  20  * http://www.gnu.org/copyleft/gpl.html
  21  *
  22  * @file
  23  */
  24
  25 /**
  26  * @ingroup API
  27  */
  28 class ApiParse extends ApiBase {
  29         private $section, $text, $pstText = null;
  30
  31         public function __construct( $main, $action ) {
  32                 parent::__construct( $main, $action );
  33         }
  34
  35         public function execute() {
  36                 // The data is hot but user-dependent, like page views, so we set vary cookies
  37                 $this->getMain()->setCacheMode( 'anon-public-user-private' );
  38
  39                 // Get parameters
  40                 $params = $this->extractRequestParams();
  41                 $text = $params['text'];
  42                 $title = $params['title'];
  43                 $page = $params['page'];
  44                 $pageid = $params['pageid'];
  45                 $oldid = $params['oldid'];
  46
  47                 if ( !is_null( $page ) && ( !is_null( $text ) || $title != 'API' ) ) {
  48                         $this->dieUsage( 'The page parameter cannot be used together with the text and title parameters', 'params' );
  49                 }
  50                 $prop = array_flip( $params['prop'] );
  51
  52                 if ( isset( $params['section'] ) ) {
  53                         $this->section = $params['section'];
  54                 } else {
  55                         $this->section = false;
  56                 }
  57
  58                 // The parser needs $wgTitle to be set, apparently the
  59                 // $title parameter in Parser::parse isn't enough *sigh*
  60                 // TODO: Does this still need $wgTitle?
  61                 global $wgParser, $wgTitle, $wgLang;
  62
  63                 // Currently unnecessary, code to act as a safeguard against any change in current behaviour of uselang breaks
  64                 $oldLang = null;
  65                 if ( isset( $params['uselang'] ) && $params['uselang'] != $wgLang->getCode() ) {
  66                         $oldLang = $wgLang; // Backup wgLang
  67                         $wgLang = Language::factory( $params['uselang'] );
  68                 }
  69
  70                 $popts = ParserOptions::newFromContext( $this->getContext() );
  71                 $popts->setTidy( true );
  72                 $popts->enableLimitReport( !$params['disablepp'] );
  73
  74                 $redirValues = null;
  75
  76                 // Return result
  77                 $result = $this->getResult();
  78
  79                 if ( !is_null( $oldid ) || !is_null( $pageid ) || !is_null( $page ) ) {
  80                         if ( !is_null( $oldid ) ) {
  81                                 // Don't use the parser cache
  82                                 $rev = Revision::newFromID( $oldid );
  83                                 if ( !$rev ) {
  84                                         $this->dieUsage( "There is no revision ID $oldid", 'missingrev' );
  85                                 }
  86                                 if ( !$rev->userCan( Revision::DELETED_TEXT ) ) {
  87                                         $this->dieUsage( "You don't have permission to view deleted revisions", 'permissiondenied' );
  88                                 }
  89
  90                                 $titleObj = $rev->getTitle();
  91
  92                                 $wgTitle = $titleObj;
  93
  94                                 // If for some reason the "oldid" is actually the current revision, it may be cached
  95                                 if ( $titleObj->getLatestRevID() === intval( $oldid ) )  {
  96                                         // May get from/save to parser cache
  97                                         $p_result = $this->getParsedSectionOrText( $titleObj, $popts, $pageid,
  98                                                  isset( $prop['wikitext'] ) ) ;
  99                                 } else { // This is an old revision, so get the text differently
 100                                         $this->text = $rev->getText( Revision::FOR_THIS_USER, $this->getUser() );
 101
 102                                         if ( $this->section !== false ) {
 103                                                 $this->text = $this->getSectionText( $this->text, 'r' . $rev->getId() );
 104                                         }
 105
 106                                         // Should we save old revision parses to the parser cache?
 107                                         $p_result = $wgParser->parse( $this->text, $titleObj, $popts );
 108                                 }
 109                         } else { // Not $oldid, but $pageid or $page
 110                                 if ( $params['redirects'] ) {
 111                                         $reqParams = array(
 112                                                 'action' => 'query',
 113                                                 'redirects' => '',
 114                                         );
 115                                         if ( !is_null ( $pageid ) ) {
 116                                                 $reqParams['pageids'] = $pageid;
 117                                         } else { // $page
 118                                                 $reqParams['titles'] = $page;
 119                                         }
 120                                         $req = new FauxRequest( $reqParams );
 121                                         $main = new ApiMain( $req );
 122                                         $main->execute();
 123                                         $data = $main->getResultData();
 124                                         $redirValues = isset( $data['query']['redirects'] )
 125                                                 ? $data['query']['redirects']
 126                                                 : array();
 127                                         $to = $page;
 128                                         foreach ( (array)$redirValues as $r ) {
 129                                                 $to = $r['to'];
 130                                         }
 131                                         $titleObj = Title::newFromText( $to );
 132                                 } else {
 133                                         if ( !is_null ( $pageid ) ) {
 134                                                 $reqParams['pageids'] = $pageid;
 135                                                 $titleObj = Title::newFromID( $pageid );
 136                                         } else { // $page
 137                                                 $to = $page;
 138                                                 $titleObj = Title::newFromText( $to );
 139                                         }
 140                                 }
 141                                 if ( !is_null ( $pageid ) ) {
 142                                         if ( !$titleObj ) {
 143                                                 // Still throw nosuchpageid error if pageid was provided
 144                                                 $this->dieUsageMsg( array( 'nosuchpageid', $pageid ) );
 145                                         }
 146                                 } elseif ( !$titleObj || !$titleObj->exists() ) {
 147                                         $this->dieUsage( "The page you specified doesn't exist", 'missingtitle' );
 148                                 }
 149                                 $wgTitle = $titleObj;
 150
 151                                 if ( isset( $prop['revid'] ) ) {
 152                                         $oldid = $titleObj->getLatestRevID();
 153                                 }
 154
 155                                 // Potentially cached
 156                                 $p_result = $this->getParsedSectionOrText( $titleObj, $popts, $pageid,
 157                                          isset( $prop['wikitext'] ) ) ;
 158                         }
 159                 } else { // Not $oldid, $pageid, $page. Hence based on $text
 160
 161                         $this->text = $text;
 162                         $titleObj = Title::newFromText( $title );
 163                         if ( !$titleObj ) {
 164                                 $this->dieUsageMsg( array( 'invalidtitle', $title ) );
 165                         }
 166                         $wgTitle = $titleObj;
 167
 168                         if ( $this->section !== false ) {
 169                                 $this->text = $this->getSectionText( $this->text, $titleObj->getText() );
 170                         }
 171
 172                         if ( $params['pst'] || $params['onlypst'] ) {
 173                                 $this->pstText = $wgParser->preSaveTransform( $this->text, $titleObj, $this->getUser(), $popts );
 174                         }
 175                         if ( $params['onlypst'] ) {
 176                                 // Build a result and bail out
 177                                 $result_array = array();
 178                                 $result_array['text'] = array();
 179                                 $result->setContent( $result_array['text'], $this->pstText );
 180                                 if ( isset( $prop['wikitext'] ) ) {
 181                                         $result_array['wikitext'] = array();
 182                                         $result->setContent( $result_array['wikitext'], $this->text );
 183                                 }
 184                                 $result->addValue( null, $this->getModuleName(), $result_array );
 185                                 return;
 186                         }
 187                         // Not cached (save or load)
 188                         $p_result = $wgParser->parse( $params['pst'] ? $this->pstText : $this->text, $titleObj, $popts );
 189                 }
 190
 191                 $result_array = array();
 192
 193                 $result_array['title'] = $titleObj->getPrefixedText();
 194
 195                 if ( !is_null( $oldid ) ) {
 196                         $result_array['revid'] = intval( $oldid );
 197                 }
 198
 199                 if ( $params['redirects'] && !is_null( $redirValues ) ) {
 200                         $result_array['redirects'] = $redirValues;
 201                 }
 202
 203                 if ( isset( $prop['text'] ) ) {
 204                         $result_array['text'] = array();
 205                         $result->setContent( $result_array['text'], $p_result->getText() );
 206                 }
 207
 208                 if ( !is_null( $params['summary'] ) ) {
 209                         $result_array['parsedsummary'] = array();
 210                         $result->setContent( $result_array['parsedsummary'], Linker::formatComment( $params['summary'], $titleObj ) );
 211                 }
 212
 213                 if ( isset( $prop['langlinks'] ) ) {
 214                         $result_array['langlinks'] = $this->formatLangLinks( $p_result->getLanguageLinks() );
 215                 }
 216                 if ( isset( $prop['languageshtml'] ) ) {
 217                         $languagesHtml = $this->languagesHtml( $p_result->getLanguageLinks() );
 218                         $result_array['languageshtml'] = array();
 219                         $result->setContent( $result_array['languageshtml'], $languagesHtml );
 220                 }
 221                 if ( isset( $prop['categories'] ) ) {
 222                         $result_array['categories'] = $this->formatCategoryLinks( $p_result->getCategories() );
 223                 }
 224                 if ( isset( $prop['categorieshtml'] ) ) {
 225                         $categoriesHtml = $this->categoriesHtml( $p_result->getCategories() );
 226                         $result_array['categorieshtml'] = array();
 227                         $result->setContent( $result_array['categorieshtml'], $categoriesHtml );
 228                 }
 229                 if ( isset( $prop['links'] ) ) {
 230                         $result_array['links'] = $this->formatLinks( $p_result->getLinks() );
 231                 }
 232                 if ( isset( $prop['templates'] ) ) {
 233                         $result_array['templates'] = $this->formatLinks( $p_result->getTemplates() );
 234                 }
 235                 if ( isset( $prop['images'] ) ) {
 236                         $result_array['images'] = array_keys( $p_result->getImages() );
 237                 }
 238                 if ( isset( $prop['externallinks'] ) ) {
 239                         $result_array['externallinks'] = array_keys( $p_result->getExternalLinks() );
 240                 }
 241                 if ( isset( $prop['sections'] ) ) {
 242                         $result_array['sections'] = $p_result->getSections();
 243                 }
 244
 245                 if ( isset( $prop['displaytitle'] ) ) {
 246                         $result_array['displaytitle'] = $p_result->getDisplayTitle() ?
 247                                                         $p_result->getDisplayTitle() :
 248                                                         $titleObj->getPrefixedText();
 249                 }
 250
 251                 if ( isset( $prop['headitems'] ) || isset( $prop['headhtml'] ) ) {
 252                         $context = $this->getContext();
 253                         $context->setTitle( $titleObj );
 254                         $context->getOutput()->addParserOutputNoText( $p_result );
 255
 256                         if ( isset( $prop['headitems'] ) ) {
 257                                 $headItems = $this->formatHeadItems( $p_result->getHeadItems() );
 258
 259                                 $css = $this->formatCss( $context->getOutput()->buildCssLinksArray() );
 260
 261                                 $scripts = array( $context->getOutput()->getHeadScripts() );
 262
 263                                 $result_array['headitems'] = array_merge( $headItems, $css, $scripts );
 264                         }
 265
 266                         if ( isset( $prop['headhtml'] ) ) {
 267                                 $result_array['headhtml'] = array();
 268                                 $result->setContent( $result_array['headhtml'], $context->getOutput()->headElement( $context->getSkin() ) );
 269                         }
 270                 }
 271
 272                 if ( isset( $prop['iwlinks'] ) ) {
 273                         $result_array['iwlinks'] = $this->formatIWLinks( $p_result->getInterwikiLinks() );
 274                 }
 275
 276                 if ( isset( $prop['wikitext'] ) ) {
 277                         $result_array['wikitext'] = array();
 278                         $result->setContent( $result_array['wikitext'], $this->text );
 279                         if ( !is_null( $this->pstText ) ) {
 280                                 $result_array['psttext'] = array();
 281                                 $result->setContent( $result_array['psttext'], $this->pstText );
 282                         }
 283                 }
 284
 285                 $result_mapping = array(
 286                         'redirects' => 'r',
 287                         'langlinks' => 'll',
 288                         'categories' => 'cl',
 289                         'links' => 'pl',
 290                         'templates' => 'tl',
 291                         'images' => 'img',
 292                         'externallinks' => 'el',
 293                         'iwlinks' => 'iw',
 294                         'sections' => 's',
 295                         'headitems' => 'hi',
 296                 );
 297                 $this->setIndexedTagNames( $result_array, $result_mapping );
 298                 $result->addValue( null, $this->getModuleName(), $result_array );
 299
 300                 if ( !is_null( $oldLang ) ) {
 301                         $wgLang = $oldLang; // Reset $wgLang to $oldLang
 302                 }
 303         }
 304
 305         /**
 306          * @param $titleObj Title
 307          * @param $popts ParserOptions
 308          * @param $pageId Int
 309          * @param $getWikitext Bool
 310          * @return ParserOutput
 311          */
 312         private function getParsedSectionOrText( $titleObj, $popts, $pageId = null, $getWikitext = false ) {
 313                 global $wgParser;
 314
 315                 $page = WikiPage::factory( $titleObj );
 316
 317                 if ( $this->section !== false ) {
 318                         $this->text = $this->getSectionText( $page->getRawText(), !is_null( $pageId )
 319                                         ? 'page id ' . $pageId : $titleObj->getText() );
 320
 321                         // Not cached (save or load)
 322                         return $wgParser->parse( $this->text, $titleObj, $popts );
 323                 } else {
 324                         // Try the parser cache first
 325                         // getParserOutput will save to Parser cache if able
 326                         $pout = $page->getParserOutput( $popts );
 327                         if ( $getWikitext ) {
 328                                 $this->text = $page->getRawText();
 329                         }
 330                         return $pout;
 331                 }
 332         }
 333
 334         private function getSectionText( $text, $what ) {
 335                 global $wgParser;
 336                 // Not cached (save or load)
 337                 $text = $wgParser->getSection( $text, $this->section, false );
 338                 if ( $text === false ) {
 339                         $this->dieUsage( "There is no section {$this->section} in " . $what, 'nosuchsection' );
 340                 }
 341                 return $text;
 342         }
 343
 344         private function formatLangLinks( $links ) {
 345                 $result = array();
 346                 foreach ( $links as $link ) {
 347                         $entry = array();
 348                         $bits = explode( ':', $link, 2 );
 349                         $title = Title::newFromText( $link );
 350
 351                         $entry['lang'] = $bits[0];
 352                         if ( $title ) {
 353                                 $entry['url'] = wfExpandUrl( $title->getFullURL(), PROTO_CURRENT );
 354                         }
 355                         $this->getResult()->setContent( $entry, $bits[1] );
 356                         $result[] = $entry;
 357                 }
 358                 return $result;
 359         }
 360
 361         private function formatCategoryLinks( $links ) {
 362                 $result = array();
 363                 foreach ( $links as $link => $sortkey ) {
 364                         $entry = array();
 365                         $entry['sortkey'] = $sortkey;
 366                         $this->getResult()->setContent( $entry, $link );
 367                         $result[] = $entry;
 368                 }
 369                 return $result;
 370         }
 371
 372         private function categoriesHtml( $categories ) {
 373                 $context = $this->getContext();
 374                 $context->getOutput()->addCategoryLinks( $categories );
 375                 return $context->getSkin()->getCategories();
 376         }
 377
 378         /**
 379          * @deprecated since 1.18 No modern skin generates language links this way, please use language links
 380          *                        data to generate your own HTML.
 381          * @param $languages array
 382          * @return string
 383          */
 384         private function languagesHtml( $languages ) {
 385                 global $wgContLang, $wgHideInterlanguageLinks;
 386
 387                 if ( $wgHideInterlanguageLinks || count( $languages ) == 0 ) {
 388                         return '';
 389                 }
 390
 391                 $s = htmlspecialchars( wfMsg( 'otherlanguages' ) . wfMsg( 'colon-separator' ) );
 392
 393                 $langs = array();
 394                 foreach ( $languages as $l ) {
 395                         $nt = Title::newFromText( $l );
 396                         $text = $wgContLang->getLanguageName( $nt->getInterwiki() );
 397
 398                         $langs[] = Html::element( 'a',
 399                                 array( 'href' => $nt->getFullURL(), 'title' => $nt->getText(), 'class' => "external" ),
 400                                 $text == '' ? $l : $text );
 401                 }
 402
 403                 $s .= implode( htmlspecialchars( wfMsgExt( 'pipe-separator', 'escapenoentities' ) ), $langs );
 404
 405                 if ( $wgContLang->isRTL() ) {
 406                         $s = Html::rawElement( 'span', array( 'dir' => "LTR" ), $s );
 407                 }
 408
 409                 return $s;
 410         }
 411
 412         private function formatLinks( $links ) {
 413                 $result = array();
 414                 foreach ( $links as $ns => $nslinks ) {
 415                         foreach ( $nslinks as $title => $id ) {
 416                                 $entry = array();
 417                                 $entry['ns'] = $ns;
 418                                 $this->getResult()->setContent( $entry, Title::makeTitle( $ns, $title )->getFullText() );
 419                                 if ( $id != 0 ) {
 420                                         $entry['exists'] = '';
 421                                 }
 422                                 $result[] = $entry;
 423                         }
 424                 }
 425                 return $result;
 426         }
 427
 428         private function formatIWLinks( $iw ) {
 429                 $result = array();
 430                 foreach ( $iw as $prefix => $titles ) {
 431                         foreach ( array_keys( $titles ) as $title ) {
 432                                 $entry = array();
 433                                 $entry['prefix'] = $prefix;
 434
 435                                 $title = Title::newFromText( "{$prefix}:{$title}" );
 436                                 if ( $title ) {
 437                                         $entry['url'] = wfExpandUrl( $title->getFullURL(), PROTO_CURRENT );
 438                                 }
 439
 440                                 $this->getResult()->setContent( $entry, $title->getFullText() );
 441                                 $result[] = $entry;
 442                         }
 443                 }
 444                 return $result;
 445         }
 446
 447         private function formatHeadItems( $headItems ) {
 448                 $result = array();
 449                 foreach ( $headItems as $tag => $content ) {
 450                         $entry = array();
 451                         $entry['tag'] = $tag;
 452                         $this->getResult()->setContent( $entry, $content );
 453                         $result[] = $entry;
 454                 }
 455                 return $result;
 456         }
 457
 458         private function formatCss( $css ) {
 459                 $result = array();
 460                 foreach ( $css as $file => $link ) {
 461                         $entry = array();
 462                         $entry['file'] = $file;
 463                         $this->getResult()->setContent( $entry, $link );
 464                         $result[] = $entry;
 465                 }
 466                 return $result;
 467         }
 468
 469         private function setIndexedTagNames( &$array, $mapping ) {
 470                 foreach ( $mapping as $key => $name ) {
 471                         if ( isset( $array[$key] ) ) {
 472                                 $this->getResult()->setIndexedTagName( $array[$key], $name );
 473                         }
 474                 }
 475         }
 476
 477         public function getAllowedParams() {
 478                 return array(
 479                         'title' => array(
 480                                 ApiBase::PARAM_DFLT => 'API',
 481                         ),
 482                         'text' => null,
 483                         'summary' => null,
 484                         'page' => null,
 485                         'pageid' => array(
 486                                 ApiBase::PARAM_TYPE => 'integer',
 487                         ),
 488                         'redirects' => false,
 489                         'oldid' => array(
 490                                 ApiBase::PARAM_TYPE => 'integer',
 491                         ),
 492                         'prop' => array(
 493                                 ApiBase::PARAM_DFLT => 'text|langlinks|categories|links|templates|images|externallinks|sections|revid|displaytitle',
 494                                 ApiBase::PARAM_ISMULTI => true,
 495                                 ApiBase::PARAM_TYPE => array(
 496                                         'text',
 497                                         'langlinks',
 498                                         'languageshtml',
 499                                         'categories',
 500                                         'categorieshtml',
 501                                         'links',
 502                                         'templates',
 503                                         'images',
 504                                         'externallinks',
 505                                         'sections',
 506                                         'revid',
 507                                         'displaytitle',
 508                                         'headitems',
 509                                         'headhtml',
 510                                         'iwlinks',
 511                                         'wikitext',
 512                                 )
 513                         ),
 514                         'pst' => false,
 515                         'onlypst' => false,
 516                         'uselang' => null,
 517                         'section' => null,
 518                         'disablepp' => false,
 519                 );
 520         }
 521
 522         public function getParamDescription() {
 523                 $p = $this->getModulePrefix();
 524                 return array(
 525                         'text' => 'Wikitext to parse',
 526                         'summary' => 'Summary to parse',
 527                         'redirects' => "If the {$p}page or the {$p}pageid parameter is set to a redirect, resolve it",
 528                         'title' => 'Title of page the text belongs to',
 529                         'page' => "Parse the content of this page. Cannot be used together with {$p}text and {$p}title",
 530                         'pageid' => "Parse the content of this page. Overrides {$p}page",
 531                         'oldid' => "Parse the content of this revision. Overrides {$p}page and {$p}pageid",
 532                         'prop' => array(
 533                                 'Which pieces of information to get',
 534                                 ' text           - Gives the parsed text of the wikitext',
 535                                 ' langlinks      - Gives the language links in the parsed wikitext',
 536                                 ' categories     - Gives the categories in the parsed wikitext',
 537                                 ' categorieshtml - Gives the HTML version of the categories',
 538                                 ' languageshtml  - Gives the HTML version of the language links',
 539                                 ' links          - Gives the internal links in the parsed wikitext',
 540                                 ' templates      - Gives the templates in the parsed wikitext',
 541                                 ' images         - Gives the images in the parsed wikitext',
 542                                 ' externallinks  - Gives the external links in the parsed wikitext',
 543                                 ' sections       - Gives the sections in the parsed wikitext',
 544                                 ' revid          - Adds the revision ID of the parsed page',
 545                                 ' displaytitle   - Adds the title of the parsed wikitext',
 546                                 ' headitems      - Gives items to put in the <head> of the page',
 547                                 ' headhtml       - Gives parsed <head> of the page',
 548                                 ' iwlinks        - Gives interwiki links in the parsed wikitext',
 549                                 ' wikitext       - Gives the original wikitext that was parsed',
 550                         ),
 551                         'pst' => array(
 552                                 'Do a pre-save transform on the input before parsing it',
 553                                 'Ignored if page, pageid or oldid is used'
 554                         ),
 555                         'onlypst' => array(
 556                                 'Do a pre-save transform (PST) on the input, but don\'t parse it',
 557                                 'Returns the same wikitext, after a PST has been applied. Ignored if page, pageid or oldid is used'
 558                         ),
 559                         'uselang' => 'Which language to parse the request in',
 560                         'section' => 'Only retrieve the content of this section number',
 561                         'disablepp' => 'Disable the PP Report from the parser output',
 562                 );
 563         }
 564
 565         public function getDescription() {
 566                 return 'Parses wikitext and returns parser output';
 567         }
 568
 569         public function getPossibleErrors() {
 570                 return array_merge( parent::getPossibleErrors(), array(
 571                         array( 'code' => 'params', 'info' => 'The page parameter cannot be used together with the text and title parameters' ),
 572                         array( 'code' => 'missingrev', 'info' => 'There is no revision ID oldid' ),
 573                         array( 'code' => 'permissiondenied', 'info' => 'You don\'t have permission to view deleted revisions' ),
 574                         array( 'code' => 'missingtitle', 'info' => 'The page you specified doesn\'t exist' ),
 575                         array( 'code' => 'nosuchsection', 'info' => 'There is no section sectionnumber in page' ),
 576                         array( 'nosuchpageid' ),
 577                         array( 'invalidtitle', 'title' ),
 578                 ) );
 579         }
 580
 581         public function getExamples() {
 582                 return array(
 583                         'api.php?action=parse&text={{Project:Sandbox}}'
 584                 );
 585         }
 586
 587         public function getHelpUrls() {
 588                 return 'http://www.mediawiki.org/wiki/API:Parsing_wikitext#parse';
 589         }
 590
 591         public function getVersion() {
 592                 return __CLASS__ . ': $Id$';
 593         }
 594 }