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         public function __construct( $main, $action ) {
  38                 parent::__construct( $main, $action );
  39         }
  40
  41         public function execute() {
  42                 // The data is hot but user-dependent, like page views, so we set vary cookies
  43                 $this->getMain()->setCacheMode( 'anon-public-user-private' );
  44
  45                 // Get parameters
  46                 $params = $this->extractRequestParams();
  47                 $text = $params['text'];
  48                 $title = $params['title'];
  49                 $page = $params['page'];
  50                 $pageid = $params['pageid'];
  51                 $oldid = $params['oldid'];
  52
  53                 if ( !is_null( $page ) && ( !is_null( $text ) || $title != 'API' ) ) {
  54                         $this->dieUsage( 'The page parameter cannot be used together with the text and title parameters', 'params' );
  55                 }
  56                 $prop = array_flip( $params['prop'] );
  57
  58                 if ( isset( $params['section'] ) ) {
  59                         $this->section = $params['section'];
  60                 } else {
  61                         $this->section = false;
  62                 }
  63
  64                 // The parser needs $wgTitle to be set, apparently the
  65                 // $title parameter in Parser::parse isn't enough *sigh*
  66                 global $wgParser, $wgUser, $wgTitle, $wgEnableParserCache, $wgLang;
  67
  68                 // Currently unnecessary, code to act as a safeguard against any change in current behaviour of uselang breaks
  69                 $oldLang = null;
  70                 if ( isset( $params['uselang'] ) && $params['uselang'] != $wgLang->getCode() ) {
  71                         $oldLang = $wgLang; // Backup wgLang
  72                         $wgLang = Language::factory( $params['uselang'] );
  73                 }
  74
  75                 $popts = new ParserOptions();
  76                 $popts->setTidy( true );
  77                 $popts->enableLimitReport( !$params['disablepp'] );
  78                 $redirValues = null;
  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                                 $text = $rev->getText( Revision::FOR_THIS_USER );
  91                                 $titleObj = $rev->getTitle();
  92                                 $wgTitle = $titleObj;
  93
  94                                 if ( $this->section !== false ) {
  95                                         $text = $this->getSectionText( $text, 'r' . $rev );
  96                                 }
  97
  98                                 $p_result = $wgParser->parse( $text, $titleObj, $popts );
  99                         } else {
 100                                 if ( !is_null ( $pageid ) ) {
 101                                         $titleObj = Title::newFromID( $pageid );
 102
 103                                         if ( !$titleObj ) {
 104                                                 $this->dieUsageMsg( array( 'nosuchpageid', $pageid ) );
 105                                         }
 106                                 } else {
 107                                         if ( $params['redirects'] ) {
 108                                                 $req = new FauxRequest( array(
 109                                                         'action' => 'query',
 110                                                         'redirects' => '',
 111                                                         'titles' => $page
 112                                                 ) );
 113                                                 $main = new ApiMain( $req );
 114                                                 $main->execute();
 115                                                 $data = $main->getResultData();
 116                                                 $redirValues = @$data['query']['redirects'];
 117                                                 $to = $page;
 118                                                 foreach ( (array)$redirValues as $r ) {
 119                                                         $to = $r['to'];
 120                                                 }
 121                                         } else {
 122                                                 $to = $page;
 123                                         }
 124                                         $titleObj = Title::newFromText( $to );
 125                                         if ( !$titleObj ) {
 126                                                 $this->dieUsage( "The page you specified doesn't exist", 'missingtitle' );
 127                                         }
 128                                 }
 129                                 $wgTitle = $titleObj;
 130
 131                                 $articleObj = new Article( $titleObj );
 132                                 if ( isset( $prop['revid'] ) ) {
 133                                         $oldid = $articleObj->getRevIdFetched();
 134                                 }
 135
 136                                 if ( $this->section !== false ) {
 137                                         $text = $this->getSectionText( $text, !is_null ( $pageid ) ? 'page id ' . $pageid : $titleObj->getText() );
 138                                         $p_result = $wgParser->parse( $text, $titleObj, $popts );
 139                                 } else {
 140                                         // Try the parser cache first
 141                                         $p_result = false;
 142                                         $pcache = ParserCache::singleton();
 143                                         if ( $wgEnableParserCache ) {
 144                                                 $p_result = $pcache->get( $articleObj, $popts );
 145                                         }
 146                                         if ( !$p_result ) {
 147                                                 $p_result = $wgParser->parse( $articleObj->getContent(), $titleObj, $popts );
 148
 149                                                 if ( $wgEnableParserCache ) {
 150                                                         $pcache->save( $p_result, $articleObj, $popts );
 151                                                 }
 152                                         }
 153                                 }
 154                         }
 155                 } else {
 156                         $titleObj = Title::newFromText( $title );
 157                         if ( !$titleObj ) {
 158                                 $titleObj = Title::newFromText( 'API' );
 159                         }
 160                         $wgTitle = $titleObj;
 161
 162                         if ( $this->section !== false ) {
 163                                 $text = $this->getSectionText( $text, $titleObj->getText() );
 164                         }
 165
 166                         if ( $params['pst'] || $params['onlypst'] ) {
 167                                 $text = $wgParser->preSaveTransform( $text, $titleObj, $wgUser, $popts );
 168                         }
 169                         if ( $params['onlypst'] ) {
 170                                 // Build a result and bail out
 171                                 $result_array['text'] = array();
 172                                 $this->getResult()->setContent( $result_array['text'], $text );
 173                                 $this->getResult()->addValue( null, $this->getModuleName(), $result_array );
 174                                 return;
 175                         }
 176                         $p_result = $wgParser->parse( $text, $titleObj, $popts );
 177                 }
 178
 179                 // Return result
 180                 $result = $this->getResult();
 181                 $result_array = array();
 182                 if ( $params['redirects'] && !is_null( $redirValues ) ) {
 183                         $result_array['redirects'] = $redirValues;
 184                 }
 185
 186                 if ( isset( $prop['text'] ) ) {
 187                         $result_array['text'] = array();
 188                         $result->setContent( $result_array['text'], $p_result->getText() );
 189                 }
 190
 191                 if ( !is_null( $params['summary'] ) ) {
 192                         $result_array['parsedsummary'] = array();
 193                         $result->setContent( $result_array['parsedsummary'], $wgUser->getSkin()->formatComment( $params['summary'], $titleObj ) );
 194                 }
 195
 196                 if ( isset( $prop['langlinks'] ) ) {
 197                         $result_array['langlinks'] = $this->formatLangLinks( $p_result->getLanguageLinks() );
 198                 }
 199                 if ( isset( $prop['categories'] ) ) {
 200                         $result_array['categories'] = $this->formatCategoryLinks( $p_result->getCategories() );
 201                 }
 202                 if ( isset( $prop['links'] ) ) {
 203                         $result_array['links'] = $this->formatLinks( $p_result->getLinks() );
 204                 }
 205                 if ( isset( $prop['templates'] ) ) {
 206                         $result_array['templates'] = $this->formatLinks( $p_result->getTemplates() );
 207                 }
 208                 if ( isset( $prop['images'] ) ) {
 209                         $result_array['images'] = array_keys( $p_result->getImages() );
 210                 }
 211                 if ( isset( $prop['externallinks'] ) ) {
 212                         $result_array['externallinks'] = array_keys( $p_result->getExternalLinks() );
 213                 }
 214                 if ( isset( $prop['sections'] ) ) {
 215                         $result_array['sections'] = $p_result->getSections();
 216                 }
 217
 218                 if ( isset( $prop['displaytitle'] ) ) {
 219                         $result_array['displaytitle'] = $p_result->getDisplayTitle() ?
 220                                                         $p_result->getDisplayTitle() :
 221                                                         $titleObj->getPrefixedText();
 222                 }
 223
 224                 if ( isset( $prop['headitems'] ) || isset( $prop['headhtml'] ) ) {
 225                         $out = new OutputPage;
 226                         $out->addParserOutputNoText( $p_result );
 227                         $userSkin = $wgUser->getSkin();
 228                 }
 229
 230                 if ( isset( $prop['headitems'] ) ) {
 231                         $headItems = $this->formatHeadItems( $p_result->getHeadItems() );
 232
 233                         $userSkin->setupUserCss( $out );
 234                         $css = $this->formatCss( $out->buildCssLinksArray() );
 235
 236                         $scripts = array( $out->getHeadScripts( $userSkin ) );
 237
 238                         $result_array['headitems'] = array_merge( $headItems, $css, $scripts );
 239                 }
 240
 241                 if ( isset( $prop['headhtml'] ) ) {
 242                         $result_array['headhtml'] = array();
 243                         $result->setContent( $result_array['headhtml'], $out->headElement( $userSkin ) );
 244                 }
 245
 246                 if ( isset( $prop['iwlinks'] ) ) {
 247                         $result_array['iwlinks'] = $this->formatIWLinks( $p_result->getInterwikiLinks() );
 248                 }
 249
 250                 if ( !is_null( $oldid ) ) {
 251                         $result_array['revid'] = intval( $oldid );
 252                 }
 253
 254                 $result_mapping = array(
 255                         'redirects' => 'r',
 256                         'langlinks' => 'll',
 257                         'categories' => 'cl',
 258                         'links' => 'pl',
 259                         'templates' => 'tl',
 260                         'images' => 'img',
 261                         'externallinks' => 'el',
 262                         'iwlinks' => 'iw',
 263                         'sections' => 's',
 264                         'headitems' => 'hi',
 265                 );
 266                 $this->setIndexedTagNames( $result_array, $result_mapping );
 267                 $result->addValue( null, $this->getModuleName(), $result_array );
 268
 269                 if ( !is_null( $oldLang ) ) {
 270                         $wgLang = $oldLang; // Reset $wgLang to $oldLang
 271                 }
 272         }
 273
 274         private function getSectionText( $text, $what ) {
 275                 global $wgParser;
 276                 $text = $wgParser->getSection( $text, $this->section, false );
 277                 if ( $text === false ) {
 278                         $this->dieUsage( "There is no section {$this->section} in " . $what, 'nosuchsection' );
 279                 }
 280                 return $text;
 281         }
 282
 283         private function formatLangLinks( $links ) {
 284                 $result = array();
 285                 foreach ( $links as $link ) {
 286                         $entry = array();
 287                         $bits = explode( ':', $link, 2 );
 288                         $entry['lang'] = $bits[0];
 289                         $this->getResult()->setContent( $entry, $bits[1] );
 290                         $result[] = $entry;
 291                 }
 292                 return $result;
 293         }
 294
 295         private function formatCategoryLinks( $links ) {
 296                 $result = array();
 297                 foreach ( $links as $link => $sortkey ) {
 298                         $entry = array();
 299                         $entry['sortkey'] = $sortkey;
 300                         $this->getResult()->setContent( $entry, $link );
 301                         $result[] = $entry;
 302                 }
 303                 return $result;
 304         }
 305
 306         private function formatLinks( $links ) {
 307                 $result = array();
 308                 foreach ( $links as $ns => $nslinks ) {
 309                         foreach ( $nslinks as $title => $id ) {
 310                                 $entry = array();
 311                                 $entry['ns'] = $ns;
 312                                 $this->getResult()->setContent( $entry, Title::makeTitle( $ns, $title )->getFullText() );
 313                                 if ( $id != 0 ) {
 314                                         $entry['exists'] = '';
 315                                 }
 316                                 $result[] = $entry;
 317                         }
 318                 }
 319                 return $result;
 320         }
 321
 322         private function formatIWLinks( $iw ) {
 323                 $result = array();
 324                 foreach ( $iw as $prefix => $titles ) {
 325                         foreach ( $titles as $title => $id ) {
 326                                 $entry = array();
 327                                 $entry['prefix'] = $prefix;
 328
 329                                 $title = Title::newFromText( "{$prefix}:{$title}" );
 330                                 if ( $title ) {
 331                                         $entry['url'] = $title->getFullURL();
 332                                 }
 333
 334                                 $this->getResult()->setContent( $entry, $title->getFullText() );
 335                                 $result[] = $entry;
 336                         }
 337                 }
 338                 return $result;
 339         }
 340
 341         private function formatHeadItems( $headItems ) {
 342                 $result = array();
 343                 foreach ( $headItems as $tag => $content ) {
 344                         $entry = array();
 345                         $entry['tag'] = $tag;
 346                         $this->getResult()->setContent( $entry, $content );
 347                         $result[] = $entry;
 348                 }
 349                 return $result;
 350         }
 351
 352         private function formatCss( $css ) {
 353                 $result = array();
 354                 foreach ( $css as $file => $link ) {
 355                         $entry = array();
 356                         $entry['file'] = $file;
 357                         $this->getResult()->setContent( $entry, $link );
 358                         $result[] = $entry;
 359                 }
 360                 return $result;
 361         }
 362
 363         private function setIndexedTagNames( &$array, $mapping ) {
 364                 foreach ( $mapping as $key => $name ) {
 365                         if ( isset( $array[$key] ) ) {
 366                                 $this->getResult()->setIndexedTagName( $array[$key], $name );
 367                         }
 368                 }
 369         }
 370
 371         public function getAllowedParams() {
 372                 return array(
 373                         'title' => array(
 374                                 ApiBase::PARAM_DFLT => 'API',
 375                         ),
 376                         'text' => null,
 377                         'summary' => null,
 378                         'page' => null,
 379                         'pageid' => null,
 380                         'redirects' => false,
 381                         'oldid' => null,
 382                         'prop' => array(
 383                                 ApiBase::PARAM_DFLT => 'text|langlinks|categories|links|templates|images|externallinks|sections|revid|displaytitle',
 384                                 ApiBase::PARAM_ISMULTI => true,
 385                                 ApiBase::PARAM_TYPE => array(
 386                                         'text',
 387                                         'langlinks',
 388                                         'categories',
 389                                         'links',
 390                                         'templates',
 391                                         'images',
 392                                         'externallinks',
 393                                         'sections',
 394                                         'revid',
 395                                         'displaytitle',
 396                                         'headitems',
 397                                         'headhtml',
 398                                         'iwlinks',
 399                                 )
 400                         ),
 401                         'pst' => false,
 402                         'onlypst' => false,
 403                         'uselang' => null,
 404                         'section' => null,
 405                         'disablepp' => false,
 406                 );
 407         }
 408
 409         public function getParamDescription() {
 410                 $p = $this->getModulePrefix();
 411                 return array(
 412                         'text' => 'Wikitext to parse',
 413                         'summary' => 'Summary to parse',
 414                         'redirects' => "If the {$p}page parameter is set to a redirect, resolve it",
 415                         'title' => 'Title of page the text belongs to',
 416                         'page' => "Parse the content of this page. Cannot be used together with {$p}text and {$p}title",
 417                         'pageid' => "Parse the content of this page. Overrides {$p}page",
 418                         'oldid' => "Parse the content of this revision. Overrides {$p}page and {$p}pageid",
 419                         'prop' => array(
 420                                 'Which pieces of information to get',
 421                                 ' text           - Gives the parsed text of the wikitext',
 422                                 ' langlinks      - Gives the langlinks the parsed wikitext',
 423                                 ' categories     - Gives the categories of the parsed wikitext',
 424                                 ' links          - Gives the internal links in the parsed wikitext',
 425                                 ' templates      - Gives the templates in the parsed wikitext',
 426                                 ' images         - Gives the images in the parsed wikitext',
 427                                 ' externallinks  - Gives the external links in the parsed wikitext',
 428                                 ' sections       - Gives the sections in the parsed wikitext',
 429                                 ' revid          - Adds the revision id of the parsed page',
 430                                 ' displaytitle   - Adds the title of the parsed wikitext',
 431                                 ' headitems      - Gives items to put in the <head> of the page',
 432                                 ' headhtml       - Gives parsed <head> of the page',
 433                                 ' iwlinks        - Gives interwiki links in the parsed wikitext',
 434                                 'NOTE: Section tree is only generated if there are more than 4 sections, or if the __TOC__ keyword is present'
 435                         ),
 436                         'pst' => array(
 437                                 'Do a pre-save transform on the input before parsing it',
 438                                 'Ignored if page, pageid or oldid is used'
 439                         ),
 440                         'onlypst' => array(
 441                                 'Do a pre-save transform (PST) on the input, but don\'t parse it',
 442                                 'Returns the same wikitext, after a PST has been applied. Ignored if page, pageid or oldid is used'
 443                         ),
 444                         'uselang' => 'Which language to parse the request in',
 445                         'section' => 'Only retrieve the content of this section number',
 446                         'disablepp' => 'Disable the PP Report from the parser output',
 447                 );
 448         }
 449
 450         public function getDescription() {
 451                 return 'This module parses wikitext and returns parser output';
 452         }
 453
 454         public function getPossibleErrors() {
 455                 return array_merge( parent::getPossibleErrors(), array(
 456                         array( 'code' => 'params', 'info' => 'The page parameter cannot be used together with the text and title parameters' ),
 457                         array( 'code' => 'missingrev', 'info' => 'There is no revision ID oldid' ),
 458                         array( 'code' => 'permissiondenied', 'info' => 'You don\'t have permission to view deleted revisions' ),
 459                         array( 'code' => 'missingtitle', 'info' => 'The page you specified doesn\'t exist' ),
 460                         array( 'code' => 'nosuchsection', 'info' => 'There is no section sectionnumber in page' ),
 461                         array( 'nosuchpageid' ),
 462                 ) );
 463         }
 464
 465         protected function getExamples() {
 466                 return array(
 467                         'api.php?action=parse&text={{Project:Sandbox}}'
 468                 );
 469         }
 470
 471         public function getVersion() {
 472                 return __CLASS__ . ': $Id$';
 473         }
 474 }