3 * Created on Dec 01, 2007
5 * Copyright © 2007 Yuri Astrakhan "<Firstname><Lastname>@gmail.com"
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.
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.
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
28 class ApiParse
extends ApiBase
{
29 private $section, $text, $pstText = null;
31 public function __construct( $main, $action ) {
32 parent
::__construct( $main, $action );
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' );
40 $params = $this->extractRequestParams();
41 $text = $params['text'];
42 $title = $params['title'];
43 $page = $params['page'];
44 $pageid = $params['pageid'];
45 $oldid = $params['oldid'];
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' );
51 $prop = array_flip( $params['prop'] );
53 if ( isset( $params['section'] ) ) {
54 $this->section
= $params['section'];
56 $this->section
= false;
59 // The parser needs $wgTitle to be set, apparently the
60 // $title parameter in Parser::parse isn't enough *sigh*
61 // TODO: Does this still need $wgTitle?
62 global $wgParser, $wgTitle;
64 // Currently unnecessary, code to act as a safeguard against any change in current behaviour of uselang breaks
66 if ( isset( $params['uselang'] ) && $params['uselang'] != $this->getContext()->getLanguage()->getCode() ) {
67 $oldLang = $this->getContext()->getLanguage(); // Backup language
68 $this->getContext()->setLanguage( Language
::factory( $params['uselang'] ) );
74 $result = $this->getResult();
76 if ( !is_null( $oldid ) ||
!is_null( $pageid ) ||
!is_null( $page ) ) {
77 if ( !is_null( $oldid ) ) {
78 // Don't use the parser cache
79 $rev = Revision
::newFromID( $oldid );
81 $this->dieUsage( "There is no revision ID $oldid", 'missingrev' );
83 if ( !$rev->userCan( Revision
::DELETED_TEXT
, $this->getUser() ) ) {
84 $this->dieUsage( "You don't have permission to view deleted revisions", 'permissiondenied' );
87 $titleObj = $rev->getTitle();
89 $pageObj = WikiPage
::factory( $titleObj );
90 $popts = $pageObj->makeParserOptions( $this->getContext() );
91 $popts->enableLimitReport( !$params['disablepp'] );
93 // If for some reason the "oldid" is actually the current revision, it may be cached
94 if ( $rev->isCurrent() ) {
95 // May get from/save to parser cache
96 $p_result = $this->getParsedSectionOrText( $pageObj, $popts, $pageid,
97 isset( $prop['wikitext'] ) ) ;
98 } else { // This is an old revision, so get the text differently
99 $this->text
= $rev->getText( Revision
::FOR_THIS_USER
, $this->getUser() );
101 if ( $this->section
!== false ) {
102 $this->text
= $this->getSectionText( $this->text
, 'r' . $rev->getId() );
105 // Should we save old revision parses to the parser cache?
106 $p_result = $wgParser->parse( $this->text
, $titleObj, $popts );
108 } else { // Not $oldid, but $pageid or $page
109 if ( $params['redirects'] ) {
114 if ( !is_null ( $pageid ) ) {
115 $reqParams['pageids'] = $pageid;
117 $reqParams['titles'] = $page;
119 $req = new FauxRequest( $reqParams );
120 $main = new ApiMain( $req );
122 $data = $main->getResultData();
123 $redirValues = isset( $data['query']['redirects'] )
124 ?
$data['query']['redirects']
127 foreach ( (array)$redirValues as $r ) {
130 $pageParams = array( 'title' => $to );
131 } elseif ( !is_null( $pageid ) ) {
132 $pageParams = array( 'pageid' => $pageid );
134 $pageParams = array( 'title' => $page );
137 $pageObj = $this->getTitleOrPageId( $pageParams, 'fromdb' );
138 $titleObj = $pageObj->getTitle();
139 $wgTitle = $titleObj;
141 if ( isset( $prop['revid'] ) ) {
142 $oldid = $pageObj->getLatest();
145 $popts = $pageObj->makeParserOptions( $this->getContext() );
146 $popts->enableLimitReport( !$params['disablepp'] );
148 // Potentially cached
149 $p_result = $this->getParsedSectionOrText( $pageObj, $popts, $pageid,
150 isset( $prop['wikitext'] ) ) ;
152 } else { // Not $oldid, $pageid, $page. Hence based on $text
154 if ( is_null( $text ) ) {
155 $this->dieUsage( 'The text parameter should be passed with the title parameter. Should you be using the "page" parameter instead?', 'params' );
158 $titleObj = Title
::newFromText( $title );
160 $this->dieUsageMsg( array( 'invalidtitle', $title ) );
162 $wgTitle = $titleObj;
163 $pageObj = WikiPage
::factory( $titleObj );
165 $popts = $pageObj->makeParserOptions( $this->getContext() );
166 $popts->enableLimitReport( !$params['disablepp'] );
168 if ( $this->section
!== false ) {
169 $this->text
= $this->getSectionText( $this->text
, $titleObj->getText() );
172 if ( $params['pst'] ||
$params['onlypst'] ) {
173 $this->pstText
= $wgParser->preSaveTransform( $this->text
, $titleObj, $this->getUser(), $popts );
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
);
184 $result->addValue( null, $this->getModuleName(), $result_array );
187 // Not cached (save or load)
188 $p_result = $wgParser->parse( $params['pst'] ?
$this->pstText
: $this->text
, $titleObj, $popts );
191 $result_array = array();
193 $result_array['title'] = $titleObj->getPrefixedText();
195 if ( !is_null( $oldid ) ) {
196 $result_array['revid'] = intval( $oldid );
199 if ( $params['redirects'] && !is_null( $redirValues ) ) {
200 $result_array['redirects'] = $redirValues;
203 if ( isset( $prop['text'] ) ) {
204 $result_array['text'] = array();
205 $result->setContent( $result_array['text'], $p_result->getText() );
208 if ( !is_null( $params['summary'] ) ) {
209 $result_array['parsedsummary'] = array();
210 $result->setContent( $result_array['parsedsummary'], Linker
::formatComment( $params['summary'], $titleObj ) );
213 if ( isset( $prop['langlinks'] ) ) {
214 $result_array['langlinks'] = $this->formatLangLinks( $p_result->getLanguageLinks() );
216 if ( isset( $prop['languageshtml'] ) ) {
217 $languagesHtml = $this->languagesHtml( $p_result->getLanguageLinks() );
218 $result_array['languageshtml'] = array();
219 $result->setContent( $result_array['languageshtml'], $languagesHtml );
221 if ( isset( $prop['categories'] ) ) {
222 $result_array['categories'] = $this->formatCategoryLinks( $p_result->getCategories() );
224 if ( isset( $prop['categorieshtml'] ) ) {
225 $categoriesHtml = $this->categoriesHtml( $p_result->getCategories() );
226 $result_array['categorieshtml'] = array();
227 $result->setContent( $result_array['categorieshtml'], $categoriesHtml );
229 if ( isset( $prop['links'] ) ) {
230 $result_array['links'] = $this->formatLinks( $p_result->getLinks() );
232 if ( isset( $prop['templates'] ) ) {
233 $result_array['templates'] = $this->formatLinks( $p_result->getTemplates() );
235 if ( isset( $prop['images'] ) ) {
236 $result_array['images'] = array_keys( $p_result->getImages() );
238 if ( isset( $prop['externallinks'] ) ) {
239 $result_array['externallinks'] = array_keys( $p_result->getExternalLinks() );
241 if ( isset( $prop['sections'] ) ) {
242 $result_array['sections'] = $p_result->getSections();
245 if ( isset( $prop['displaytitle'] ) ) {
246 $result_array['displaytitle'] = $p_result->getDisplayTitle() ?
247 $p_result->getDisplayTitle() :
248 $titleObj->getPrefixedText();
251 if ( isset( $prop['headitems'] ) ||
isset( $prop['headhtml'] ) ) {
252 $context = $this->getContext();
253 $context->setTitle( $titleObj );
254 $context->getOutput()->addParserOutputNoText( $p_result );
256 if ( isset( $prop['headitems'] ) ) {
257 $headItems = $this->formatHeadItems( $p_result->getHeadItems() );
259 $css = $this->formatCss( $context->getOutput()->buildCssLinksArray() );
261 $scripts = array( $context->getOutput()->getHeadScripts() );
263 $result_array['headitems'] = array_merge( $headItems, $css, $scripts );
266 if ( isset( $prop['headhtml'] ) ) {
267 $result_array['headhtml'] = array();
268 $result->setContent( $result_array['headhtml'], $context->getOutput()->headElement( $context->getSkin() ) );
272 if ( isset( $prop['iwlinks'] ) ) {
273 $result_array['iwlinks'] = $this->formatIWLinks( $p_result->getInterwikiLinks() );
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
);
284 if ( isset( $prop['properties'] ) ) {
285 $result_array['properties'] = $this->formatProperties( $p_result->getProperties() );
288 if ( $params['generatexml'] ) {
289 $wgParser->startExternalParse( $titleObj, $popts, OT_PREPROCESS
);
290 $dom = $wgParser->preprocessToDom( $this->text
);
291 if ( is_callable( array( $dom, 'saveXML' ) ) ) {
292 $xml = $dom->saveXML();
294 $xml = $dom->__toString();
296 $result_array['parsetree'] = array();
297 $result->setContent( $result_array['parsetree'], $xml );
300 $result_mapping = array(
303 'categories' => 'cl',
307 'externallinks' => 'el',
311 'properties' => 'pp',
313 $this->setIndexedTagNames( $result_array, $result_mapping );
314 $result->addValue( null, $this->getModuleName(), $result_array );
316 if ( !is_null( $oldLang ) ) {
317 $this->getContext()->setLanguage( $oldLang ); // Reset language to $oldLang
322 * @param $page WikiPage
323 * @param $popts ParserOptions
325 * @param $getWikitext Bool
326 * @return ParserOutput
328 private function getParsedSectionOrText( $page, $popts, $pageId = null, $getWikitext = false ) {
331 if ( $this->section
!== false ) {
332 $this->text
= $this->getSectionText( $page->getRawText(), !is_null( $pageId )
333 ?
'page id ' . $pageId : $page->getTitle()->getPrefixedText() );
335 // Not cached (save or load)
336 return $wgParser->parse( $this->text
, $page->getTitle(), $popts );
338 // Try the parser cache first
339 // getParserOutput will save to Parser cache if able
340 $pout = $page->getParserOutput( $popts );
342 $this->dieUsage( "There is no revision ID {$page->getLatest()}", 'missingrev' );
344 if ( $getWikitext ) {
345 $this->text
= $page->getRawText();
351 private function getSectionText( $text, $what ) {
353 // Not cached (save or load)
354 $text = $wgParser->getSection( $text, $this->section
, false );
355 if ( $text === false ) {
356 $this->dieUsage( "There is no section {$this->section} in " . $what, 'nosuchsection' );
361 private function formatLangLinks( $links ) {
363 foreach ( $links as $link ) {
365 $bits = explode( ':', $link, 2 );
366 $title = Title
::newFromText( $link );
368 $entry['lang'] = $bits[0];
370 $entry['url'] = wfExpandUrl( $title->getFullURL(), PROTO_CURRENT
);
372 $this->getResult()->setContent( $entry, $bits[1] );
378 private function formatCategoryLinks( $links ) {
380 foreach ( $links as $link => $sortkey ) {
382 $entry['sortkey'] = $sortkey;
383 $this->getResult()->setContent( $entry, $link );
389 private function categoriesHtml( $categories ) {
390 $context = $this->getContext();
391 $context->getOutput()->addCategoryLinks( $categories );
392 return $context->getSkin()->getCategories();
396 * @deprecated since 1.18 No modern skin generates language links this way, please use language links
397 * data to generate your own HTML.
398 * @param $languages array
401 private function languagesHtml( $languages ) {
402 wfDeprecated( __METHOD__
, '1.18' );
404 global $wgContLang, $wgHideInterlanguageLinks;
406 if ( $wgHideInterlanguageLinks ||
count( $languages ) == 0 ) {
410 $s = htmlspecialchars( wfMessage( 'otherlanguages' )->text() . wfMessage( 'colon-separator' )->text() );
413 foreach ( $languages as $l ) {
414 $nt = Title
::newFromText( $l );
415 $text = Language
::fetchLanguageName( $nt->getInterwiki() );
417 $langs[] = Html
::element( 'a',
418 array( 'href' => $nt->getFullURL(), 'title' => $nt->getText(), 'class' => "external" ),
419 $text == '' ?
$l : $text );
422 $s .= implode( wfMessage( 'pipe-separator' )->escaped(), $langs );
424 if ( $wgContLang->isRTL() ) {
425 $s = Html
::rawElement( 'span', array( 'dir' => "LTR" ), $s );
431 private function formatLinks( $links ) {
433 foreach ( $links as $ns => $nslinks ) {
434 foreach ( $nslinks as $title => $id ) {
437 $this->getResult()->setContent( $entry, Title
::makeTitle( $ns, $title )->getFullText() );
439 $entry['exists'] = '';
447 private function formatIWLinks( $iw ) {
449 foreach ( $iw as $prefix => $titles ) {
450 foreach ( array_keys( $titles ) as $title ) {
452 $entry['prefix'] = $prefix;
454 $title = Title
::newFromText( "{$prefix}:{$title}" );
456 $entry['url'] = wfExpandUrl( $title->getFullURL(), PROTO_CURRENT
);
459 $this->getResult()->setContent( $entry, $title->getFullText() );
466 private function formatHeadItems( $headItems ) {
468 foreach ( $headItems as $tag => $content ) {
470 $entry['tag'] = $tag;
471 $this->getResult()->setContent( $entry, $content );
477 private function formatProperties( $properties ) {
479 foreach ( $properties as $name => $value ) {
481 $entry['name'] = $name;
482 $this->getResult()->setContent( $entry, $value );
488 private function formatCss( $css ) {
490 foreach ( $css as $file => $link ) {
492 $entry['file'] = $file;
493 $this->getResult()->setContent( $entry, $link );
499 private function setIndexedTagNames( &$array, $mapping ) {
500 foreach ( $mapping as $key => $name ) {
501 if ( isset( $array[$key] ) ) {
502 $this->getResult()->setIndexedTagName( $array[$key], $name );
507 public function getAllowedParams() {
510 ApiBase
::PARAM_DFLT
=> 'API',
516 ApiBase
::PARAM_TYPE
=> 'integer',
518 'redirects' => false,
520 ApiBase
::PARAM_TYPE
=> 'integer',
523 ApiBase
::PARAM_DFLT
=> 'text|langlinks|categories|links|templates|images|externallinks|sections|revid|displaytitle|iwlinks|properties',
524 ApiBase
::PARAM_ISMULTI
=> true,
525 ApiBase
::PARAM_TYPE
=> array(
549 'disablepp' => false,
550 'generatexml' => false,
554 public function getParamDescription() {
555 $p = $this->getModulePrefix();
557 'text' => 'Wikitext to parse',
558 'summary' => 'Summary to parse',
559 'redirects' => "If the {$p}page or the {$p}pageid parameter is set to a redirect, resolve it",
560 'title' => 'Title of page the text belongs to',
561 'page' => "Parse the content of this page. Cannot be used together with {$p}text and {$p}title",
562 'pageid' => "Parse the content of this page. Overrides {$p}page",
563 'oldid' => "Parse the content of this revision. Overrides {$p}page and {$p}pageid",
565 'Which pieces of information to get',
566 ' text - Gives the parsed text of the wikitext',
567 ' langlinks - Gives the language links in the parsed wikitext',
568 ' categories - Gives the categories in the parsed wikitext',
569 ' categorieshtml - Gives the HTML version of the categories',
570 ' languageshtml - Gives the HTML version of the language links',
571 ' links - Gives the internal links in the parsed wikitext',
572 ' templates - Gives the templates in the parsed wikitext',
573 ' images - Gives the images in the parsed wikitext',
574 ' externallinks - Gives the external links in the parsed wikitext',
575 ' sections - Gives the sections in the parsed wikitext',
576 ' revid - Adds the revision ID of the parsed page',
577 ' displaytitle - Adds the title of the parsed wikitext',
578 ' headitems - Gives items to put in the <head> of the page',
579 ' headhtml - Gives parsed <head> of the page',
580 ' iwlinks - Gives interwiki links in the parsed wikitext',
581 ' wikitext - Gives the original wikitext that was parsed',
582 ' properties - Gives various properties defined in the parsed wikitext',
585 'Do a pre-save transform on the input before parsing it',
586 'Ignored if page, pageid or oldid is used'
589 'Do a pre-save transform (PST) on the input, but don\'t parse it',
590 'Returns the same wikitext, after a PST has been applied. Ignored if page, pageid or oldid is used'
592 'uselang' => 'Which language to parse the request in',
593 'section' => 'Only retrieve the content of this section number',
594 'disablepp' => 'Disable the PP Report from the parser output',
595 'generatexml' => 'Generate XML parse tree',
599 public function getDescription() {
601 'Parses wikitext and returns parser output',
602 'See the various prop-Modules of action=query to get information from the current version of a page',
606 public function getPossibleErrors() {
607 return array_merge( parent
::getPossibleErrors(), array(
608 array( 'code' => 'params', 'info' => 'The page parameter cannot be used together with the text and title parameters' ),
609 array( 'code' => 'params', 'info' => 'The text parameter should be passed with the title parameter. Should you be using the "page" parameter instead?' ),
610 array( 'code' => 'missingrev', 'info' => 'There is no revision ID oldid' ),
611 array( 'code' => 'permissiondenied', 'info' => 'You don\'t have permission to view deleted revisions' ),
612 array( 'code' => 'missingtitle', 'info' => 'The page you specified doesn\'t exist' ),
613 array( 'code' => 'nosuchsection', 'info' => 'There is no section sectionnumber in page' ),
614 array( 'nosuchpageid' ),
615 array( 'invalidtitle', 'title' ),
619 public function getExamples() {
621 'api.php?action=parse&text={{Project:Sandbox}}'
625 public function getHelpUrls() {
626 return 'https://www.mediawiki.org/wiki/API:Parsing_wikitext#parse';
629 public function getVersion() {
630 return __CLASS__
. ': $Id$';