3 * File for Parser and related classes
9 * Update this version number when the ParserOutput format
10 * changes in an incompatible way, so the parser cache
11 * can automatically discard old data.
13 define( 'MW_PARSER_VERSION', '1.6.1' );
15 define( 'RLH_FOR_UPDATE', 1 );
17 # Allowed values for $mOutputType
18 define( 'OT_HTML', 1 );
19 define( 'OT_WIKI', 2 );
20 define( 'OT_MSG' , 3 );
21 define( 'OT_PREPROCESS', 4 );
23 # Flags for setFunctionHook
24 define( 'SFH_NO_HASH', 1 );
26 # string parameter for extractTags which will cause it
27 # to strip HTML comments in addition to regular
28 # <XML>-style tags. This should not be anything we
29 # may want to use in wikisyntax
30 define( 'STRIP_COMMENTS', 'HTMLCommentStrip' );
32 # Constants needed for external link processing
33 define( 'HTTP_PROTOCOLS', 'http:\/\/|https:\/\/' );
34 # Everything except bracket, space, or control characters
35 define( 'EXT_LINK_URL_CLASS', '[^][<>"\\x00-\\x20\\x7F]' );
36 # Including space, but excluding newlines
37 define( 'EXT_LINK_TEXT_CLASS', '[^\]\\x0a\\x0d]' );
38 define( 'EXT_IMAGE_FNAME_CLASS', '[A-Za-z0-9_.,~%\\-+&;#*?!=()@\\x80-\\xFF]' );
39 define( 'EXT_IMAGE_EXTENSIONS', 'gif|png|jpg|jpeg' );
40 define( 'EXT_LINK_BRACKETED', '/\[(\b(' . wfUrlProtocols() . ')'.
41 EXT_LINK_URL_CLASS
.'+) *('.EXT_LINK_TEXT_CLASS
.'*?)\]/S' );
42 define( 'EXT_IMAGE_REGEX',
43 '/^('.HTTP_PROTOCOLS
.')'. # Protocol
44 '('.EXT_LINK_URL_CLASS
.'+)\\/'. # Hostname and path
45 '('.EXT_IMAGE_FNAME_CLASS
.'+)\\.((?i)'.EXT_IMAGE_EXTENSIONS
.')$/S' # Filename
48 // State constants for the definition list colon extraction
49 define( 'MW_COLON_STATE_TEXT', 0 );
50 define( 'MW_COLON_STATE_TAG', 1 );
51 define( 'MW_COLON_STATE_TAGSTART', 2 );
52 define( 'MW_COLON_STATE_CLOSETAG', 3 );
53 define( 'MW_COLON_STATE_TAGSLASH', 4 );
54 define( 'MW_COLON_STATE_COMMENT', 5 );
55 define( 'MW_COLON_STATE_COMMENTDASH', 6 );
56 define( 'MW_COLON_STATE_COMMENTDASHDASH', 7 );
59 * PHP Parser - Processes wiki markup (which uses a more user-friendly
60 * syntax, such as "[[link]]" for making links), and provides a one-way
61 * transformation of that wiki markup it into XHTML output / markup
62 * (which in turn the browser understands, and can display).
65 * There are four main entry points into the Parser class:
67 * produces HTML output
69 * produces altered wiki markup.
71 * performs brace substitution on MediaWiki messages
73 * removes HTML comments and expands templates
76 * objects: $wgLang, $wgContLang
78 * NOT $wgArticle, $wgUser or $wgTitle. Keep them away!
81 * $wgUseTex*, $wgUseDynamicDates*, $wgInterwikiMagic*,
82 * $wgNamespacesWithSubpages, $wgAllowExternalImages*,
83 * $wgLocaltimezone, $wgAllowSpecialInclusion*,
86 * * only within ParserOptions
93 const VERSION
= MW_PARSER_VERSION
;
98 var $mTagHooks, $mFunctionHooks, $mFunctionSynonyms, $mVariables;
100 # Cleared with clearState():
101 var $mOutput, $mAutonumber, $mDTopen, $mStripState;
102 var $mIncludeCount, $mArgStack, $mLastSection, $mInPre;
103 var $mInterwikiLinkHolders, $mLinkHolders, $mUniqPrefix;
104 var $mIncludeSizes, $mDefaultSort;
105 var $mTemplates, // cache of already loaded templates, avoids
106 // multiple SQL queries for the same string
107 $mTemplatePath; // stores an unsorted hash of all the templates already loaded
108 // in this path. Used for loop detection.
111 # These are variables reset at least once per parse regardless of $clearState
112 var $mOptions, // ParserOptions object
113 $mTitle, // Title context, used for self-link rendering and similar things
114 $mOutputType, // Output type, one of the OT_xxx constants
115 $ot, // Shortcut alias, see setOutputType()
116 $mRevisionId, // ID to display in {{REVISIONID}} tags
117 $mRevisionTimestamp, // The timestamp of the specified revision ID
118 $mRevIdForTs; // The revision ID which was used to fetch the timestamp
128 $this->mTagHooks
= array();
129 $this->mFunctionHooks
= array();
130 $this->mFunctionSynonyms
= array( 0 => array(), 1 => array() );
131 $this->mFirstCall
= true;
135 * Do various kinds of initialisation on the first call of the parser
137 function firstCallInit() {
138 if ( !$this->mFirstCall
) {
142 wfProfileIn( __METHOD__
);
143 global $wgAllowDisplayTitle, $wgAllowSlowParserFunctions;
145 $this->setHook( 'pre', array( $this, 'renderPreTag' ) );
147 $this->setFunctionHook( 'int', array( 'CoreParserFunctions', 'intFunction' ), SFH_NO_HASH
);
148 $this->setFunctionHook( 'ns', array( 'CoreParserFunctions', 'ns' ), SFH_NO_HASH
);
149 $this->setFunctionHook( 'urlencode', array( 'CoreParserFunctions', 'urlencode' ), SFH_NO_HASH
);
150 $this->setFunctionHook( 'lcfirst', array( 'CoreParserFunctions', 'lcfirst' ), SFH_NO_HASH
);
151 $this->setFunctionHook( 'ucfirst', array( 'CoreParserFunctions', 'ucfirst' ), SFH_NO_HASH
);
152 $this->setFunctionHook( 'lc', array( 'CoreParserFunctions', 'lc' ), SFH_NO_HASH
);
153 $this->setFunctionHook( 'uc', array( 'CoreParserFunctions', 'uc' ), SFH_NO_HASH
);
154 $this->setFunctionHook( 'localurl', array( 'CoreParserFunctions', 'localurl' ), SFH_NO_HASH
);
155 $this->setFunctionHook( 'localurle', array( 'CoreParserFunctions', 'localurle' ), SFH_NO_HASH
);
156 $this->setFunctionHook( 'fullurl', array( 'CoreParserFunctions', 'fullurl' ), SFH_NO_HASH
);
157 $this->setFunctionHook( 'fullurle', array( 'CoreParserFunctions', 'fullurle' ), SFH_NO_HASH
);
158 $this->setFunctionHook( 'formatnum', array( 'CoreParserFunctions', 'formatnum' ), SFH_NO_HASH
);
159 $this->setFunctionHook( 'grammar', array( 'CoreParserFunctions', 'grammar' ), SFH_NO_HASH
);
160 $this->setFunctionHook( 'plural', array( 'CoreParserFunctions', 'plural' ), SFH_NO_HASH
);
161 $this->setFunctionHook( 'numberofpages', array( 'CoreParserFunctions', 'numberofpages' ), SFH_NO_HASH
);
162 $this->setFunctionHook( 'numberofusers', array( 'CoreParserFunctions', 'numberofusers' ), SFH_NO_HASH
);
163 $this->setFunctionHook( 'numberofarticles', array( 'CoreParserFunctions', 'numberofarticles' ), SFH_NO_HASH
);
164 $this->setFunctionHook( 'numberoffiles', array( 'CoreParserFunctions', 'numberoffiles' ), SFH_NO_HASH
);
165 $this->setFunctionHook( 'numberofadmins', array( 'CoreParserFunctions', 'numberofadmins' ), SFH_NO_HASH
);
166 $this->setFunctionHook( 'numberofedits', array( 'CoreParserFunctions', 'numberofedits' ), SFH_NO_HASH
);
167 $this->setFunctionHook( 'language', array( 'CoreParserFunctions', 'language' ), SFH_NO_HASH
);
168 $this->setFunctionHook( 'padleft', array( 'CoreParserFunctions', 'padleft' ), SFH_NO_HASH
);
169 $this->setFunctionHook( 'padright', array( 'CoreParserFunctions', 'padright' ), SFH_NO_HASH
);
170 $this->setFunctionHook( 'anchorencode', array( 'CoreParserFunctions', 'anchorencode' ), SFH_NO_HASH
);
171 $this->setFunctionHook( 'special', array( 'CoreParserFunctions', 'special' ) );
172 $this->setFunctionHook( 'defaultsort', array( 'CoreParserFunctions', 'defaultsort' ), SFH_NO_HASH
);
174 if ( $wgAllowDisplayTitle ) {
175 $this->setFunctionHook( 'displaytitle', array( 'CoreParserFunctions', 'displaytitle' ), SFH_NO_HASH
);
177 if ( $wgAllowSlowParserFunctions ) {
178 $this->setFunctionHook( 'pagesinnamespace', array( 'CoreParserFunctions', 'pagesinnamespace' ), SFH_NO_HASH
);
181 $this->initialiseVariables();
182 $this->mFirstCall
= false;
183 wfProfileOut( __METHOD__
);
191 function clearState() {
192 wfProfileIn( __METHOD__
);
193 if ( $this->mFirstCall
) {
194 $this->firstCallInit();
196 $this->mOutput
= new ParserOutput
;
197 $this->mAutonumber
= 0;
198 $this->mLastSection
= '';
199 $this->mDTopen
= false;
200 $this->mIncludeCount
= array();
201 $this->mStripState
= new StripState
;
202 $this->mArgStack
= array();
203 $this->mInPre
= false;
204 $this->mInterwikiLinkHolders
= array(
208 $this->mLinkHolders
= array(
209 'namespaces' => array(),
211 'queries' => array(),
215 $this->mRevisionTimestamp
= $this->mRevisionId
= null;
218 * Prefix for temporary replacement strings for the multipass parser.
219 * \x07 should never appear in input as it's disallowed in XML.
220 * Using it at the front also gives us a little extra robustness
221 * since it shouldn't match when butted up against identifier-like
224 $this->mUniqPrefix
= "\x07UNIQ" . Parser
::getRandomString();
226 # Clear these on every parse, bug 4549
227 $this->mTemplates
= array();
228 $this->mTemplatePath
= array();
230 $this->mShowToc
= true;
231 $this->mForceTocPosition
= false;
232 $this->mIncludeSizes
= array(
237 $this->mDefaultSort
= false;
239 wfRunHooks( 'ParserClearState', array( &$this ) );
240 wfProfileOut( __METHOD__
);
243 function setOutputType( $ot ) {
244 $this->mOutputType
= $ot;
247 'html' => $ot == OT_HTML
,
248 'wiki' => $ot == OT_WIKI
,
249 'msg' => $ot == OT_MSG
,
250 'pre' => $ot == OT_PREPROCESS
,
255 * Accessor for mUniqPrefix.
259 function uniqPrefix() {
260 return $this->mUniqPrefix
;
264 * Convert wikitext to HTML
265 * Do not call this function recursively.
267 * @param string $text Text we want to parse
268 * @param Title &$title A title object
269 * @param array $options
270 * @param boolean $linestart
271 * @param boolean $clearState
272 * @param int $revid number to pass in {{REVISIONID}}
273 * @return ParserOutput a ParserOutput
275 public function parse( $text, &$title, $options, $linestart = true, $clearState = true, $revid = null ) {
277 * First pass--just handle <nowiki> sections, pass the rest off
278 * to internalParse() which does all the real work.
281 global $wgUseTidy, $wgAlwaysUseTidy, $wgContLang;
282 $fname = 'Parser::parse-' . wfGetCaller();
283 wfProfileIn( __METHOD__
);
284 wfProfileIn( $fname );
290 $this->mOptions
= $options;
291 $this->mTitle
=& $title;
292 $oldRevisionId = $this->mRevisionId
;
293 $oldRevisionTimestamp = $this->mRevisionTimestamp
;
294 if( $revid !== null ) {
295 $this->mRevisionId
= $revid;
296 $this->mRevisionTimestamp
= null;
298 $this->setOutputType( OT_HTML
);
299 wfRunHooks( 'ParserBeforeStrip', array( &$this, &$text, &$this->mStripState
) );
300 $text = $this->strip( $text, $this->mStripState
);
301 wfRunHooks( 'ParserAfterStrip', array( &$this, &$text, &$this->mStripState
) );
302 $text = $this->internalParse( $text );
303 $text = $this->mStripState
->unstripGeneral( $text );
305 # Clean up special characters, only run once, next-to-last before doBlockLevels
307 # french spaces, last one Guillemet-left
308 # only if there is something before the space
309 '/(.) (?=\\?|:|;|!|\\302\\273)/' => '\\1 \\2',
310 # french spaces, Guillemet-right
311 '/(\\302\\253) /' => '\\1 ',
313 $text = preg_replace( array_keys($fixtags), array_values($fixtags), $text );
316 $text = $this->doBlockLevels( $text, $linestart );
318 $this->replaceLinkHolders( $text );
320 # the position of the parserConvert() call should not be changed. it
321 # assumes that the links are all replaced and the only thing left
322 # is the <nowiki> mark.
323 # Side-effects: this calls $this->mOutput->setTitleText()
324 $text = $wgContLang->parserConvert( $text, $this );
326 $text = $this->mStripState
->unstripNoWiki( $text );
328 wfRunHooks( 'ParserBeforeTidy', array( &$this, &$text ) );
330 $text = Sanitizer
::normalizeCharReferences( $text );
332 if (($wgUseTidy and $this->mOptions
->mTidy
) or $wgAlwaysUseTidy) {
333 $text = Parser
::tidy($text);
335 # attempt to sanitize at least some nesting problems
336 # (bug #2702 and quite a few others)
338 # ''Something [http://www.cool.com cool''] -->
339 # <i>Something</i><a href="http://www.cool.com"..><i>cool></i></a>
340 '/(<([bi])>)(<([bi])>)?([^<]*)(<\/?a[^<]*>)([^<]*)(<\/\\4>)?(<\/\\2>)/' =>
341 '\\1\\3\\5\\8\\9\\6\\1\\3\\7\\8\\9',
342 # fix up an anchor inside another anchor, only
343 # at least for a single single nested link (bug 3695)
344 '/(<a[^>]+>)([^<]*)(<a[^>]+>[^<]*)<\/a>(.*)<\/a>/' =>
345 '\\1\\2</a>\\3</a>\\1\\4</a>',
346 # fix div inside inline elements- doBlockLevels won't wrap a line which
347 # contains a div, so fix it up here; replace
348 # div with escaped text
349 '/(<([aib]) [^>]+>)([^<]*)(<div([^>]*)>)(.*)(<\/div>)([^<]*)(<\/\\2>)/' =>
350 '\\1\\3<div\\5>\\6</div>\\8\\9',
351 # remove empty italic or bold tag pairs, some
352 # introduced by rules above
353 '/<([bi])><\/\\1>/' => '',
356 $text = preg_replace(
357 array_keys( $tidyregs ),
358 array_values( $tidyregs ),
362 wfRunHooks( 'ParserAfterTidy', array( &$this, &$text ) );
364 # Information on include size limits, for the benefit of users who try to skirt them
365 if ( max( $this->mIncludeSizes
) > 1000 ) {
366 $max = $this->mOptions
->getMaxIncludeSize();
368 "Pre-expand include size: {$this->mIncludeSizes['pre-expand']} bytes\n" .
369 "Post-expand include size: {$this->mIncludeSizes['post-expand']} bytes\n" .
370 "Template argument size: {$this->mIncludeSizes['arg']} bytes\n" .
371 "Maximum: $max bytes\n" .
374 $this->mOutput
->setText( $text );
375 $this->mRevisionId
= $oldRevisionId;
376 $this->mRevisionTimestamp
= $oldRevisionTimestamp;
377 wfProfileOut( $fname );
378 wfProfileOut( __METHOD__
);
380 return $this->mOutput
;
384 * Recursive parser entry point that can be called from an extension tag
387 function recursiveTagParse( $text ) {
388 wfProfileIn( __METHOD__
);
389 wfRunHooks( 'ParserBeforeStrip', array( &$this, &$text, &$this->mStripState
) );
390 $text = $this->strip( $text, $this->mStripState
);
391 wfRunHooks( 'ParserAfterStrip', array( &$this, &$text, &$this->mStripState
) );
392 $text = $this->internalParse( $text );
393 wfProfileOut( __METHOD__
);
398 * Expand templates and variables in the text, producing valid, static wikitext.
399 * Also removes comments.
401 function preprocess( $text, $title, $options ) {
402 wfProfileIn( __METHOD__
);
404 $this->setOutputType( OT_PREPROCESS
);
405 $this->mOptions
= $options;
406 $this->mTitle
= $title;
407 wfRunHooks( 'ParserBeforeStrip', array( &$this, &$text, &$this->mStripState
) );
408 $text = $this->strip( $text, $this->mStripState
);
409 wfRunHooks( 'ParserAfterStrip', array( &$this, &$text, &$this->mStripState
) );
410 if ( $this->mOptions
->getRemoveComments() ) {
411 $text = Sanitizer
::removeHTMLcomments( $text );
413 $text = $this->replaceVariables( $text );
414 $text = $this->mStripState
->unstripBoth( $text );
415 wfProfileOut( __METHOD__
);
420 * Get a random string
425 function getRandomString() {
426 return dechex(mt_rand(0, 0x7fffffff)) . dechex(mt_rand(0, 0x7fffffff));
429 function &getTitle() { return $this->mTitle
; }
430 function getOptions() { return $this->mOptions
; }
432 function getFunctionLang() {
433 global $wgLang, $wgContLang;
434 return $this->mOptions
->getInterfaceMessage() ?
$wgLang : $wgContLang;
438 * Replaces all occurrences of HTML-style comments and the given tags
439 * in the text with a random marker and returns teh next text. The output
440 * parameter $matches will be an associative array filled with data in
442 * 'UNIQ-xxxxx' => array(
445 * array( 'param' => 'x' ),
446 * '<element param="x">tag content</element>' ) )
448 * @param $elements list of element names. Comments are always extracted.
449 * @param $text Source text string.
450 * @param $uniq_prefix
455 function extractTagsAndParams($elements, $text, &$matches, $uniq_prefix = ''){
460 $taglist = implode( '|', $elements );
461 $start = "/<($taglist)(\\s+[^>]*?|\\s*?)(\/?>)|<(!--)/i";
463 while ( '' != $text ) {
464 $p = preg_split( $start, $text, 2, PREG_SPLIT_DELIM_CAPTURE
);
466 if( count( $p ) < 5 ) {
469 if( count( $p ) > 5 ) {
483 $marker = "$uniq_prefix-$element-" . sprintf('%08X', $n++
) . '-QINU';
484 $stripped .= $marker;
486 if ( $close === '/>' ) {
487 // Empty element tag, <tag />
492 if( $element == '!--' ) {
495 $end = "/(<\\/$element\\s*>)/i";
497 $q = preg_split( $end, $inside, 2, PREG_SPLIT_DELIM_CAPTURE
);
499 if( count( $q ) < 3 ) {
500 # No end tag -- let it run out to the end of the text.
509 $matches[$marker] = array( $element,
511 Sanitizer
::decodeTagAttributes( $attributes ),
512 "<$element$attributes$close$content$tail" );
518 * Strips and renders nowiki, pre, math, hiero
519 * If $render is set, performs necessary rendering operations on plugins
520 * Returns the text, and fills an array with data needed in unstrip()
522 * @param StripState $state
524 * @param bool $stripcomments when set, HTML comments <!-- like this -->
525 * will be stripped in addition to other tags. This is important
526 * for section editing, where these comments cause confusion when
527 * counting the sections in the wikisource
529 * @param array dontstrip contains tags which should not be stripped;
530 * used to prevent stipping of <gallery> when saving (fixes bug 2700)
534 function strip( $text, $state, $stripcomments = false , $dontstrip = array () ) {
536 wfProfileIn( __METHOD__
);
537 $render = ($this->mOutputType
== OT_HTML
);
539 $uniq_prefix = $this->mUniqPrefix
;
540 $commentState = new ReplacementArray
;
541 $nowikiItems = array();
542 $generalItems = array();
544 $elements = array_merge(
545 array( 'nowiki', 'gallery' ),
546 array_keys( $this->mTagHooks
) );
549 $elements[] = 'html';
551 if( $this->mOptions
->getUseTeX() ) {
552 $elements[] = 'math';
555 # Removing $dontstrip tags from $elements list (currently only 'gallery', fixing bug 2700)
556 foreach ( $elements AS $k => $v ) {
557 if ( !in_array ( $v , $dontstrip ) ) continue;
558 unset ( $elements[$k] );
562 $text = Parser
::extractTagsAndParams( $elements, $text, $matches, $uniq_prefix );
564 foreach( $matches as $marker => $data ) {
565 list( $element, $content, $params, $tag ) = $data;
567 $tagName = strtolower( $element );
568 wfProfileIn( __METHOD__
."-render-$tagName" );
572 if( substr( $tag, -3 ) == '-->' ) {
575 // Unclosed comment in input.
576 // Close it so later stripping can remove it
585 // Shouldn't happen otherwise. :)
587 $output = Xml
::escapeTagsOnly( $content );
590 $output = $wgContLang->armourMath( MathRenderer
::renderMath( $content ) );
593 $output = $this->renderImageGallery( $content, $params );
596 if( isset( $this->mTagHooks
[$tagName] ) ) {
597 $output = call_user_func_array( $this->mTagHooks
[$tagName],
598 array( $content, $params, $this ) );
600 throw new MWException( "Invalid call hook $element" );
603 wfProfileOut( __METHOD__
."-render-$tagName" );
605 // Just stripping tags; keep the source
609 // Unstrip the output, to support recursive strip() calls
610 $output = $state->unstripBoth( $output );
612 if( !$stripcomments && $element == '!--' ) {
613 $commentState->setPair( $marker, $output );
614 } elseif ( $element == 'html' ||
$element == 'nowiki' ) {
615 $nowikiItems[$marker] = $output;
617 $generalItems[$marker] = $output;
620 # Add the new items to the state
621 # We do this after the loop instead of during it to avoid slowing
622 # down the recursive unstrip
623 $state->nowiki
->mergeArray( $nowikiItems );
624 $state->general
->mergeArray( $generalItems );
626 # Unstrip comments unless explicitly told otherwise.
627 # (The comments are always stripped prior to this point, so as to
628 # not invoke any extension tags / parser hooks contained within
630 if ( !$stripcomments ) {
631 // Put them all back and forget them
632 $text = $commentState->replace( $text );
635 wfProfileOut( __METHOD__
);
640 * Restores pre, math, and other extensions removed by strip()
642 * always call unstripNoWiki() after this one
644 * @deprecated use $this->mStripState->unstrip()
646 function unstrip( $text, $state ) {
647 return $state->unstripGeneral( $text );
651 * Always call this after unstrip() to preserve the order
654 * @deprecated use $this->mStripState->unstrip()
656 function unstripNoWiki( $text, $state ) {
657 return $state->unstripNoWiki( $text );
661 * @deprecated use $this->mStripState->unstripBoth()
663 function unstripForHTML( $text ) {
664 return $this->mStripState
->unstripBoth( $text );
668 * Add an item to the strip state
669 * Returns the unique tag which must be inserted into the stripped text
670 * The tag will be replaced with the original text in unstrip()
674 function insertStripItem( $text, &$state ) {
675 $rnd = $this->mUniqPrefix
. '-item' . Parser
::getRandomString();
676 $state->general
->setPair( $rnd, $text );
681 * Interface with html tidy, used if $wgUseTidy = true.
682 * If tidy isn't able to correct the markup, the original will be
683 * returned in all its glory with a warning comment appended.
685 * Either the external tidy program or the in-process tidy extension
686 * will be used depending on availability. Override the default
687 * $wgTidyInternal setting to disable the internal if it's not working.
689 * @param string $text Hideous HTML input
690 * @return string Corrected HTML output
694 function tidy( $text ) {
695 global $wgTidyInternal;
696 $wrappedtext = '<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN"'.
697 ' "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"><html>'.
698 '<head><title>test</title></head><body>'.$text.'</body></html>';
699 if( $wgTidyInternal ) {
700 $correctedtext = Parser
::internalTidy( $wrappedtext );
702 $correctedtext = Parser
::externalTidy( $wrappedtext );
704 if( is_null( $correctedtext ) ) {
705 wfDebug( "Tidy error detected!\n" );
706 return $text . "\n<!-- Tidy found serious XHTML errors -->\n";
708 return $correctedtext;
712 * Spawn an external HTML tidy process and get corrected markup back from it.
717 function externalTidy( $text ) {
718 global $wgTidyConf, $wgTidyBin, $wgTidyOpts;
719 $fname = 'Parser::externalTidy';
720 wfProfileIn( $fname );
725 $descriptorspec = array(
726 0 => array('pipe', 'r'),
727 1 => array('pipe', 'w'),
728 2 => array('file', '/dev/null', 'a') // FIXME: this line in UNIX-specific, it generates a warning on Windows, because /dev/null is not a valid Windows file.
731 $process = proc_open("$wgTidyBin -config $wgTidyConf $wgTidyOpts$opts", $descriptorspec, $pipes);
732 if (is_resource($process)) {
733 // Theoretically, this style of communication could cause a deadlock
734 // here. If the stdout buffer fills up, then writes to stdin could
735 // block. This doesn't appear to happen with tidy, because tidy only
736 // writes to stdout after it's finished reading from stdin. Search
737 // for tidyParseStdin and tidySaveStdout in console/tidy.c
738 fwrite($pipes[0], $text);
740 while (!feof($pipes[1])) {
741 $cleansource .= fgets($pipes[1], 1024);
744 proc_close($process);
747 wfProfileOut( $fname );
749 if( $cleansource == '' && $text != '') {
750 // Some kind of error happened, so we couldn't get the corrected text.
751 // Just give up; we'll use the source text and append a warning.
759 * Use the HTML tidy PECL extension to use the tidy library in-process,
760 * saving the overhead of spawning a new process. Currently written to
761 * the PHP 4.3.x version of the extension, may not work on PHP 5.
763 * 'pear install tidy' should be able to compile the extension module.
768 function internalTidy( $text ) {
770 $fname = 'Parser::internalTidy';
771 wfProfileIn( $fname );
773 tidy_load_config( $wgTidyConf );
774 tidy_set_encoding( 'utf8' );
775 tidy_parse_string( $text );
777 if( tidy_get_status() == 2 ) {
778 // 2 is magic number for fatal error
779 // http://www.php.net/manual/en/function.tidy-get-status.php
782 $cleansource = tidy_get_output();
784 wfProfileOut( $fname );
789 * parse the wiki syntax used to render tables
793 function doTableStuff ( $text ) {
794 $fname = 'Parser::doTableStuff';
795 wfProfileIn( $fname );
797 $lines = explode ( "\n" , $text );
798 $td_history = array (); // Is currently a td tag open?
799 $last_tag_history = array (); // Save history of last lag activated (td, th or caption)
800 $tr_history = array (); // Is currently a tr tag open?
801 $tr_attributes = array (); // history of tr attributes
802 $has_opened_tr = array(); // Did this table open a <tr> element?
803 $indent_level = 0; // indent level of the table
804 foreach ( $lines as $key => $line )
806 $line = trim ( $line );
808 if( $line == '' ) { // empty line, go to next line
811 $first_character = $line{0};
814 if ( preg_match( '/^(:*)\{\|(.*)$/' , $line , $matches ) ) {
815 // First check if we are starting a new table
816 $indent_level = strlen( $matches[1] );
818 $attributes = $this->mStripState
->unstripBoth( $matches[2] );
819 $attributes = Sanitizer
::fixTagAttributes ( $attributes , 'table' );
821 $lines[$key] = str_repeat( '<dl><dd>' , $indent_level ) . "<table{$attributes}>";
822 array_push ( $td_history , false );
823 array_push ( $last_tag_history , '' );
824 array_push ( $tr_history , false );
825 array_push ( $tr_attributes , '' );
826 array_push ( $has_opened_tr , false );
827 } else if ( count ( $td_history ) == 0 ) {
828 // Don't do any of the following
830 } else if ( substr ( $line , 0 , 2 ) == '|}' ) {
831 // We are ending a table
832 $line = '</table>' . substr ( $line , 2 );
833 $last_tag = array_pop ( $last_tag_history );
835 if ( !array_pop ( $has_opened_tr ) ) {
836 $line = "<tr><td></td></tr>{$line}";
839 if ( array_pop ( $tr_history ) ) {
840 $line = "</tr>{$line}";
843 if ( array_pop ( $td_history ) ) {
844 $line = "</{$last_tag}>{$line}";
846 array_pop ( $tr_attributes );
847 $lines[$key] = $line . str_repeat( '</dd></dl>' , $indent_level );
848 } else if ( substr ( $line , 0 , 2 ) == '|-' ) {
849 // Now we have a table row
850 $line = preg_replace( '#^\|-+#', '', $line );
852 // Whats after the tag is now only attributes
853 $attributes = $this->mStripState
->unstripBoth( $line );
854 $attributes = Sanitizer
::fixTagAttributes ( $attributes , 'tr' );
855 array_pop ( $tr_attributes );
856 array_push ( $tr_attributes , $attributes );
859 $last_tag = array_pop ( $last_tag_history );
860 array_pop ( $has_opened_tr );
861 array_push ( $has_opened_tr , true );
863 if ( array_pop ( $tr_history ) ) {
867 if ( array_pop ( $td_history ) ) {
868 $line = "</{$last_tag}>{$line}";
871 $lines[$key] = $line;
872 array_push ( $tr_history , false );
873 array_push ( $td_history , false );
874 array_push ( $last_tag_history , '' );
876 else if ( $first_character == '|' ||
$first_character == '!' ||
substr ( $line , 0 , 2 ) == '|+' ) {
877 // This might be cell elements, td, th or captions
878 if ( substr ( $line , 0 , 2 ) == '|+' ) {
879 $first_character = '+';
880 $line = substr ( $line , 1 );
883 $line = substr ( $line , 1 );
885 if ( $first_character == '!' ) {
886 $line = str_replace ( '!!' , '||' , $line );
889 // Split up multiple cells on the same line.
890 // FIXME : This can result in improper nesting of tags processed
891 // by earlier parser steps, but should avoid splitting up eg
892 // attribute values containing literal "||".
893 $cells = StringUtils
::explodeMarkup( '||' , $line );
897 // Loop through each table cell
898 foreach ( $cells as $cell )
901 if ( $first_character != '+' )
903 $tr_after = array_pop ( $tr_attributes );
904 if ( !array_pop ( $tr_history ) ) {
905 $previous = "<tr{$tr_after}>\n";
907 array_push ( $tr_history , true );
908 array_push ( $tr_attributes , '' );
909 array_pop ( $has_opened_tr );
910 array_push ( $has_opened_tr , true );
913 $last_tag = array_pop ( $last_tag_history );
915 if ( array_pop ( $td_history ) ) {
916 $previous = "</{$last_tag}>{$previous}";
919 if ( $first_character == '|' ) {
921 } else if ( $first_character == '!' ) {
923 } else if ( $first_character == '+' ) {
924 $last_tag = 'caption';
929 array_push ( $last_tag_history , $last_tag );
931 // A cell could contain both parameters and data
932 $cell_data = explode ( '|' , $cell , 2 );
934 // Bug 553: Note that a '|' inside an invalid link should not
935 // be mistaken as delimiting cell parameters
936 if ( strpos( $cell_data[0], '[[' ) !== false ) {
937 $cell = "{$previous}<{$last_tag}>{$cell}";
938 } else if ( count ( $cell_data ) == 1 )
939 $cell = "{$previous}<{$last_tag}>{$cell_data[0]}";
941 $attributes = $this->mStripState
->unstripBoth( $cell_data[0] );
942 $attributes = Sanitizer
::fixTagAttributes( $attributes , $last_tag );
943 $cell = "{$previous}<{$last_tag}{$attributes}>{$cell_data[1]}";
946 $lines[$key] .= $cell;
947 array_push ( $td_history , true );
952 // Closing open td, tr && table
953 while ( count ( $td_history ) > 0 )
955 if ( array_pop ( $td_history ) ) {
958 if ( array_pop ( $tr_history ) ) {
961 if ( !array_pop ( $has_opened_tr ) ) {
962 $lines[] = "<tr><td></td></tr>" ;
965 $lines[] = '</table>' ;
968 $output = implode ( "\n" , $lines ) ;
970 // special case: don't return empty table
971 if( $output == "<table>\n<tr><td></td></tr>\n</table>" ) {
975 wfProfileOut( $fname );
981 * Helper function for parse() that transforms wiki markup into
982 * HTML. Only called for $mOutputType == OT_HTML.
986 function internalParse( $text ) {
989 $fname = 'Parser::internalParse';
990 wfProfileIn( $fname );
992 # Hook to suspend the parser in this state
993 if ( !wfRunHooks( 'ParserBeforeInternalParse', array( &$this, &$text, &$this->mStripState
) ) ) {
994 wfProfileOut( $fname );
998 # Remove <noinclude> tags and <includeonly> sections
999 $text = strtr( $text, array( '<onlyinclude>' => '' , '</onlyinclude>' => '' ) );
1000 $text = strtr( $text, array( '<noinclude>' => '', '</noinclude>' => '') );
1001 $text = StringUtils
::delimiterReplace( '<includeonly>', '</includeonly>', '', $text );
1003 $text = Sanitizer
::removeHTMLtags( $text, array( &$this, 'attributeStripCallback' ) );
1005 $text = $this->replaceVariables( $text, $args );
1006 wfRunHooks( 'InternalParseBeforeLinks', array( &$this, &$text, &$this->mStripState
) );
1008 // Tables need to come after variable replacement for things to work
1009 // properly; putting them before other transformations should keep
1010 // exciting things like link expansions from showing up in surprising
1012 $text = $this->doTableStuff( $text );
1014 $text = preg_replace( '/(^|\n)-----*/', '\\1<hr />', $text );
1016 $text = $this->stripToc( $text );
1017 $this->stripNoGallery( $text );
1018 $text = $this->doHeadings( $text );
1019 if($this->mOptions
->getUseDynamicDates()) {
1020 $df =& DateFormatter
::getInstance();
1021 $text = $df->reformat( $this->mOptions
->getDateFormat(), $text );
1023 $text = $this->doAllQuotes( $text );
1024 $text = $this->replaceInternalLinks( $text );
1025 $text = $this->replaceExternalLinks( $text );
1027 # replaceInternalLinks may sometimes leave behind
1028 # absolute URLs, which have to be masked to hide them from replaceExternalLinks
1029 $text = str_replace($this->mUniqPrefix
."NOPARSE", "", $text);
1031 $text = $this->doMagicLinks( $text );
1032 $text = $this->formatHeadings( $text, $isMain );
1034 wfProfileOut( $fname );
1039 * Replace special strings like "ISBN xxx" and "RFC xxx" with
1040 * magic external links.
1044 function &doMagicLinks( &$text ) {
1045 wfProfileIn( __METHOD__
);
1046 $text = preg_replace_callback(
1048 <a.*?</a> | # Skip link text
1049 <.*?> | # Skip stuff inside HTML elements
1050 (?:RFC|PMID)\s+([0-9]+) | # RFC or PMID, capture number as m[1]
1051 ISBN\s+(\b # ISBN, capture number as m[2]
1052 (?: 97[89] [\ \-]? )? # optional 13-digit ISBN prefix
1053 (?: [0-9] [\ \-]? ){9} # 9 digits with opt. delimiters
1054 [0-9Xx] # check digit
1056 )!x', array( &$this, 'magicLinkCallback' ), $text );
1057 wfProfileOut( __METHOD__
);
1061 function magicLinkCallback( $m ) {
1062 if ( substr( $m[0], 0, 1 ) == '<' ) {
1065 } elseif ( substr( $m[0], 0, 4 ) == 'ISBN' ) {
1067 $num = strtr( $isbn, array(
1072 $titleObj = SpecialPage
::getTitleFor( 'Booksources' );
1073 $text = '<a href="' .
1074 $titleObj->escapeLocalUrl( "isbn=$num" ) .
1075 "\" class=\"internal\">ISBN $isbn</a>";
1077 if ( substr( $m[0], 0, 3 ) == 'RFC' ) {
1081 } elseif ( substr( $m[0], 0, 4 ) == 'PMID' ) {
1083 $urlmsg = 'pubmedurl';
1086 throw new MWException( __METHOD__
.': unrecognised match type "' .
1087 substr($m[0], 0, 20 ) . '"' );
1090 $url = wfMsg( $urlmsg, $id);
1091 $sk = $this->mOptions
->getSkin();
1092 $la = $sk->getExternalLinkAttributes( $url, $keyword.$id );
1093 $text = "<a href=\"{$url}\"{$la}>{$keyword} {$id}</a>";
1099 * Parse headers and return html
1103 function doHeadings( $text ) {
1104 $fname = 'Parser::doHeadings';
1105 wfProfileIn( $fname );
1106 for ( $i = 6; $i >= 1; --$i ) {
1107 $h = str_repeat( '=', $i );
1108 $text = preg_replace( "/^{$h}(.+){$h}\\s*$/m",
1109 "<h{$i}>\\1</h{$i}>\\2", $text );
1111 wfProfileOut( $fname );
1116 * Replace single quotes with HTML markup
1118 * @return string the altered text
1120 function doAllQuotes( $text ) {
1121 $fname = 'Parser::doAllQuotes';
1122 wfProfileIn( $fname );
1124 $lines = explode( "\n", $text );
1125 foreach ( $lines as $line ) {
1126 $outtext .= $this->doQuotes ( $line ) . "\n";
1128 $outtext = substr($outtext, 0,-1);
1129 wfProfileOut( $fname );
1134 * Helper function for doAllQuotes()
1137 function doQuotes( $text ) {
1138 $arr = preg_split( "/(''+)/", $text, -1, PREG_SPLIT_DELIM_CAPTURE
);
1139 if ( count( $arr ) == 1 )
1143 # First, do some preliminary work. This may shift some apostrophes from
1144 # being mark-up to being text. It also counts the number of occurrences
1145 # of bold and italics mark-ups.
1149 foreach ( $arr as $r )
1151 if ( ( $i %
2 ) == 1 )
1153 # If there are ever four apostrophes, assume the first is supposed to
1154 # be text, and the remaining three constitute mark-up for bold text.
1155 if ( strlen( $arr[$i] ) == 4 )
1160 # If there are more than 5 apostrophes in a row, assume they're all
1161 # text except for the last 5.
1162 else if ( strlen( $arr[$i] ) > 5 )
1164 $arr[$i-1] .= str_repeat( "'", strlen( $arr[$i] ) - 5 );
1167 # Count the number of occurrences of bold and italics mark-ups.
1168 # We are not counting sequences of five apostrophes.
1169 if ( strlen( $arr[$i] ) == 2 ) { $numitalics++
; }
1170 else if ( strlen( $arr[$i] ) == 3 ) { $numbold++
; }
1171 else if ( strlen( $arr[$i] ) == 5 ) { $numitalics++
; $numbold++
; }
1176 # If there is an odd number of both bold and italics, it is likely
1177 # that one of the bold ones was meant to be an apostrophe followed
1178 # by italics. Which one we cannot know for certain, but it is more
1179 # likely to be one that has a single-letter word before it.
1180 if ( ( $numbold %
2 == 1 ) && ( $numitalics %
2 == 1 ) )
1183 $firstsingleletterword = -1;
1184 $firstmultiletterword = -1;
1186 foreach ( $arr as $r )
1188 if ( ( $i %
2 == 1 ) and ( strlen( $r ) == 3 ) )
1190 $x1 = substr ($arr[$i-1], -1);
1191 $x2 = substr ($arr[$i-1], -2, 1);
1193 if ($firstspace == -1) $firstspace = $i;
1194 } else if ($x2 == ' ') {
1195 if ($firstsingleletterword == -1) $firstsingleletterword = $i;
1197 if ($firstmultiletterword == -1) $firstmultiletterword = $i;
1203 # If there is a single-letter word, use it!
1204 if ($firstsingleletterword > -1)
1206 $arr [ $firstsingleletterword ] = "''";
1207 $arr [ $firstsingleletterword-1 ] .= "'";
1209 # If not, but there's a multi-letter word, use that one.
1210 else if ($firstmultiletterword > -1)
1212 $arr [ $firstmultiletterword ] = "''";
1213 $arr [ $firstmultiletterword-1 ] .= "'";
1215 # ... otherwise use the first one that has neither.
1216 # (notice that it is possible for all three to be -1 if, for example,
1217 # there is only one pentuple-apostrophe in the line)
1218 else if ($firstspace > -1)
1220 $arr [ $firstspace ] = "''";
1221 $arr [ $firstspace-1 ] .= "'";
1225 # Now let's actually convert our apostrophic mush to HTML!
1230 foreach ($arr as $r)
1234 if ($state == 'both')
1241 if (strlen ($r) == 2)
1244 { $output .= '</i>'; $state = ''; }
1245 else if ($state == 'bi')
1246 { $output .= '</i>'; $state = 'b'; }
1247 else if ($state == 'ib')
1248 { $output .= '</b></i><b>'; $state = 'b'; }
1249 else if ($state == 'both')
1250 { $output .= '<b><i>'.$buffer.'</i>'; $state = 'b'; }
1251 else # $state can be 'b' or ''
1252 { $output .= '<i>'; $state .= 'i'; }
1254 else if (strlen ($r) == 3)
1257 { $output .= '</b>'; $state = ''; }
1258 else if ($state == 'bi')
1259 { $output .= '</i></b><i>'; $state = 'i'; }
1260 else if ($state == 'ib')
1261 { $output .= '</b>'; $state = 'i'; }
1262 else if ($state == 'both')
1263 { $output .= '<i><b>'.$buffer.'</b>'; $state = 'i'; }
1264 else # $state can be 'i' or ''
1265 { $output .= '<b>'; $state .= 'b'; }
1267 else if (strlen ($r) == 5)
1270 { $output .= '</b><i>'; $state = 'i'; }
1271 else if ($state == 'i')
1272 { $output .= '</i><b>'; $state = 'b'; }
1273 else if ($state == 'bi')
1274 { $output .= '</i></b>'; $state = ''; }
1275 else if ($state == 'ib')
1276 { $output .= '</b></i>'; $state = ''; }
1277 else if ($state == 'both')
1278 { $output .= '<i><b>'.$buffer.'</b></i>'; $state = ''; }
1279 else # ($state == '')
1280 { $buffer = ''; $state = 'both'; }
1285 # Now close all remaining tags. Notice that the order is important.
1286 if ($state == 'b' ||
$state == 'ib')
1288 if ($state == 'i' ||
$state == 'bi' ||
$state == 'ib')
1292 # There might be lonely ''''', so make sure we have a buffer
1293 if ($state == 'both' && $buffer)
1294 $output .= '<b><i>'.$buffer.'</i></b>';
1300 * Replace external links
1302 * Note: this is all very hackish and the order of execution matters a lot.
1303 * Make sure to run maintenance/parserTests.php if you change this code.
1307 function replaceExternalLinks( $text ) {
1309 $fname = 'Parser::replaceExternalLinks';
1310 wfProfileIn( $fname );
1312 $sk = $this->mOptions
->getSkin();
1314 $bits = preg_split( EXT_LINK_BRACKETED
, $text, -1, PREG_SPLIT_DELIM_CAPTURE
);
1316 $s = $this->replaceFreeExternalLinks( array_shift( $bits ) );
1319 while ( $i<count( $bits ) ) {
1321 $protocol = $bits[$i++
];
1322 $text = $bits[$i++
];
1323 $trail = $bits[$i++
];
1325 # The characters '<' and '>' (which were escaped by
1326 # removeHTMLtags()) should not be included in
1327 # URLs, per RFC 2396.
1329 if (preg_match('/&(lt|gt);/', $url, $m2, PREG_OFFSET_CAPTURE
)) {
1330 $text = substr($url, $m2[0][1]) . ' ' . $text;
1331 $url = substr($url, 0, $m2[0][1]);
1334 # If the link text is an image URL, replace it with an <img> tag
1335 # This happened by accident in the original parser, but some people used it extensively
1336 $img = $this->maybeMakeExternalImage( $text );
1337 if ( $img !== false ) {
1343 # Set linktype for CSS - if URL==text, link is essentially free
1344 $linktype = ($text == $url) ?
'free' : 'text';
1346 # No link text, e.g. [http://domain.tld/some.link]
1347 if ( $text == '' ) {
1348 # Autonumber if allowed. See bug #5918
1349 if ( strpos( wfUrlProtocols(), substr($protocol, 0, strpos($protocol, ':')) ) !== false ) {
1350 $text = '[' . ++
$this->mAutonumber
. ']';
1351 $linktype = 'autonumber';
1353 # Otherwise just use the URL
1354 $text = htmlspecialchars( $url );
1358 # Have link text, e.g. [http://domain.tld/some.link text]s
1360 list( $dtrail, $trail ) = Linker
::splitTrail( $trail );
1363 $text = $wgContLang->markNoConversion($text);
1365 $url = Sanitizer
::cleanUrl( $url );
1367 # Process the trail (i.e. everything after this link up until start of the next link),
1368 # replacing any non-bracketed links
1369 $trail = $this->replaceFreeExternalLinks( $trail );
1371 # Use the encoded URL
1372 # This means that users can paste URLs directly into the text
1373 # Funny characters like ö aren't valid in URLs anyway
1374 # This was changed in August 2004
1375 $s .= $sk->makeExternalLink( $url, $text, false, $linktype, $this->mTitle
->getNamespace() ) . $dtrail . $trail;
1377 # Register link in the output object.
1378 # Replace unnecessary URL escape codes with the referenced character
1379 # This prevents spammers from hiding links from the filters
1380 $pasteurized = Parser
::replaceUnusualEscapes( $url );
1381 $this->mOutput
->addExternalLink( $pasteurized );
1384 wfProfileOut( $fname );
1389 * Replace anything that looks like a URL with a link
1392 function replaceFreeExternalLinks( $text ) {
1394 $fname = 'Parser::replaceFreeExternalLinks';
1395 wfProfileIn( $fname );
1397 $bits = preg_split( '/(\b(?:' . wfUrlProtocols() . '))/S', $text, -1, PREG_SPLIT_DELIM_CAPTURE
);
1398 $s = array_shift( $bits );
1401 $sk = $this->mOptions
->getSkin();
1403 while ( $i < count( $bits ) ){
1404 $protocol = $bits[$i++
];
1405 $remainder = $bits[$i++
];
1408 if ( preg_match( '/^('.EXT_LINK_URL_CLASS
.'+)(.*)$/s', $remainder, $m ) ) {
1409 # Found some characters after the protocol that look promising
1410 $url = $protocol . $m[1];
1413 # special case: handle urls as url args:
1414 # http://www.example.com/foo?=http://www.example.com/bar
1415 if(strlen($trail) == 0 &&
1417 preg_match('/^'. wfUrlProtocols() . '$/S', $bits[$i]) &&
1418 preg_match( '/^('.EXT_LINK_URL_CLASS
.'+)(.*)$/s', $bits[$i +
1], $m ))
1421 $url .= $bits[$i] . $m[1]; # protocol, url as arg to previous link
1426 # The characters '<' and '>' (which were escaped by
1427 # removeHTMLtags()) should not be included in
1428 # URLs, per RFC 2396.
1430 if (preg_match('/&(lt|gt);/', $url, $m2, PREG_OFFSET_CAPTURE
)) {
1431 $trail = substr($url, $m2[0][1]) . $trail;
1432 $url = substr($url, 0, $m2[0][1]);
1435 # Move trailing punctuation to $trail
1437 # If there is no left bracket, then consider right brackets fair game too
1438 if ( strpos( $url, '(' ) === false ) {
1442 $numSepChars = strspn( strrev( $url ), $sep );
1443 if ( $numSepChars ) {
1444 $trail = substr( $url, -$numSepChars ) . $trail;
1445 $url = substr( $url, 0, -$numSepChars );
1448 $url = Sanitizer
::cleanUrl( $url );
1450 # Is this an external image?
1451 $text = $this->maybeMakeExternalImage( $url );
1452 if ( $text === false ) {
1453 # Not an image, make a link
1454 $text = $sk->makeExternalLink( $url, $wgContLang->markNoConversion($url), true, 'free', $this->mTitle
->getNamespace() );
1455 # Register it in the output object...
1456 # Replace unnecessary URL escape codes with their equivalent characters
1457 $pasteurized = Parser
::replaceUnusualEscapes( $url );
1458 $this->mOutput
->addExternalLink( $pasteurized );
1460 $s .= $text . $trail;
1462 $s .= $protocol . $remainder;
1465 wfProfileOut( $fname );
1470 * Replace unusual URL escape codes with their equivalent characters
1474 * @todo This can merge genuinely required bits in the path or query string,
1475 * breaking legit URLs. A proper fix would treat the various parts of
1476 * the URL differently; as a workaround, just use the output for
1477 * statistical records, not for actual linking/output.
1479 static function replaceUnusualEscapes( $url ) {
1480 return preg_replace_callback( '/%[0-9A-Fa-f]{2}/',
1481 array( 'Parser', 'replaceUnusualEscapesCallback' ), $url );
1485 * Callback function used in replaceUnusualEscapes().
1486 * Replaces unusual URL escape codes with their equivalent character
1490 private static function replaceUnusualEscapesCallback( $matches ) {
1491 $char = urldecode( $matches[0] );
1492 $ord = ord( $char );
1493 // Is it an unsafe or HTTP reserved character according to RFC 1738?
1494 if ( $ord > 32 && $ord < 127 && strpos( '<>"#{}|\^~[]`;/?', $char ) === false ) {
1495 // No, shouldn't be escaped
1498 // Yes, leave it escaped
1504 * make an image if it's allowed, either through the global
1505 * option or through the exception
1508 function maybeMakeExternalImage( $url ) {
1509 $sk = $this->mOptions
->getSkin();
1510 $imagesfrom = $this->mOptions
->getAllowExternalImagesFrom();
1511 $imagesexception = !empty($imagesfrom);
1513 if ( $this->mOptions
->getAllowExternalImages()
1514 ||
( $imagesexception && strpos( $url, $imagesfrom ) === 0 ) ) {
1515 if ( preg_match( EXT_IMAGE_REGEX
, $url ) ) {
1517 $text = $sk->makeExternalImage( htmlspecialchars( $url ) );
1524 * Process [[ ]] wikilinks
1528 function replaceInternalLinks( $s ) {
1530 static $fname = 'Parser::replaceInternalLinks' ;
1532 wfProfileIn( $fname );
1534 wfProfileIn( $fname.'-setup' );
1536 # the % is needed to support urlencoded titles as well
1537 if ( !$tc ) { $tc = Title
::legalChars() . '#%'; }
1539 $sk = $this->mOptions
->getSkin();
1541 #split the entire text string on occurences of [[
1542 $a = explode( '[[', ' ' . $s );
1543 #get the first element (all text up to first [[), and remove the space we added
1544 $s = array_shift( $a );
1545 $s = substr( $s, 1 );
1547 # Match a link having the form [[namespace:link|alternate]]trail
1549 if ( !$e1 ) { $e1 = "/^([{$tc}]+)(?:\\|(.+?))?]](.*)\$/sD"; }
1550 # Match cases where there is no "]]", which might still be images
1551 static $e1_img = FALSE;
1552 if ( !$e1_img ) { $e1_img = "/^([{$tc}]+)\\|(.*)\$/sD"; }
1553 # Match the end of a line for a word that's not followed by whitespace,
1554 # e.g. in the case of 'The Arab al[[Razi]]', 'al' will be matched
1555 $e2 = wfMsgForContent( 'linkprefix' );
1557 $useLinkPrefixExtension = $wgContLang->linkPrefixExtension();
1558 if( is_null( $this->mTitle
) ) {
1559 throw new MWException( __METHOD__
.": \$this->mTitle is null\n" );
1561 $nottalk = !$this->mTitle
->isTalkPage();
1563 if ( $useLinkPrefixExtension ) {
1565 if ( preg_match( $e2, $s, $m ) ) {
1566 $first_prefix = $m[2];
1568 $first_prefix = false;
1574 if($wgContLang->hasVariants()) {
1575 $selflink = $wgContLang->convertLinkToAllVariants($this->mTitle
->getPrefixedText());
1577 $selflink = array($this->mTitle
->getPrefixedText());
1579 $useSubpages = $this->areSubpagesAllowed();
1580 wfProfileOut( $fname.'-setup' );
1582 # Loop for each link
1583 for ($k = 0; isset( $a[$k] ); $k++
) {
1585 if ( $useLinkPrefixExtension ) {
1586 wfProfileIn( $fname.'-prefixhandling' );
1587 if ( preg_match( $e2, $s, $m ) ) {
1595 $prefix = $first_prefix;
1596 $first_prefix = false;
1598 wfProfileOut( $fname.'-prefixhandling' );
1601 $might_be_img = false;
1603 wfProfileIn( "$fname-e1" );
1604 if ( preg_match( $e1, $line, $m ) ) { # page with normal text or alt
1606 # If we get a ] at the beginning of $m[3] that means we have a link that's something like:
1607 # [[Image:Foo.jpg|[http://example.com desc]]] <- having three ] in a row fucks up,
1608 # the real problem is with the $e1 regex
1611 # Still some problems for cases where the ] is meant to be outside punctuation,
1612 # and no image is in sight. See bug 2095.
1615 substr( $m[3], 0, 1 ) === ']' &&
1616 strpos($text, '[') !== false
1619 $text .= ']'; # so that replaceExternalLinks($text) works later
1620 $m[3] = substr( $m[3], 1 );
1622 # fix up urlencoded title texts
1623 if( strpos( $m[1], '%' ) !== false ) {
1624 # Should anchors '#' also be rejected?
1625 $m[1] = str_replace( array('<', '>'), array('<', '>'), urldecode($m[1]) );
1628 } elseif( preg_match($e1_img, $line, $m) ) { # Invalid, but might be an image with a link in its caption
1629 $might_be_img = true;
1631 if ( strpos( $m[1], '%' ) !== false ) {
1632 $m[1] = urldecode($m[1]);
1635 } else { # Invalid form; output directly
1636 $s .= $prefix . '[[' . $line ;
1637 wfProfileOut( "$fname-e1" );
1640 wfProfileOut( "$fname-e1" );
1641 wfProfileIn( "$fname-misc" );
1643 # Don't allow internal links to pages containing
1644 # PROTO: where PROTO is a valid URL protocol; these
1645 # should be external links.
1646 if (preg_match('/^\b(?:' . wfUrlProtocols() . ')/', $m[1])) {
1647 $s .= $prefix . '[[' . $line ;
1651 # Make subpage if necessary
1652 if( $useSubpages ) {
1653 $link = $this->maybeDoSubpageLink( $m[1], $text );
1658 $noforce = (substr($m[1], 0, 1) != ':');
1660 # Strip off leading ':'
1661 $link = substr($link, 1);
1664 wfProfileOut( "$fname-misc" );
1665 wfProfileIn( "$fname-title" );
1666 $nt = Title
::newFromText( $this->mStripState
->unstripNoWiki($link) );
1668 $s .= $prefix . '[[' . $line;
1669 wfProfileOut( "$fname-title" );
1673 $ns = $nt->getNamespace();
1674 $iw = $nt->getInterWiki();
1675 wfProfileOut( "$fname-title" );
1677 if ($might_be_img) { # if this is actually an invalid link
1678 wfProfileIn( "$fname-might_be_img" );
1679 if ($ns == NS_IMAGE
&& $noforce) { #but might be an image
1681 while (isset ($a[$k+
1]) ) {
1682 #look at the next 'line' to see if we can close it there
1683 $spliced = array_splice( $a, $k +
1, 1 );
1684 $next_line = array_shift( $spliced );
1685 $m = explode( ']]', $next_line, 3 );
1686 if ( count( $m ) == 3 ) {
1687 # the first ]] closes the inner link, the second the image
1689 $text .= "[[{$m[0]}]]{$m[1]}";
1692 } elseif ( count( $m ) == 2 ) {
1693 #if there's exactly one ]] that's fine, we'll keep looking
1694 $text .= "[[{$m[0]}]]{$m[1]}";
1696 #if $next_line is invalid too, we need look no further
1697 $text .= '[[' . $next_line;
1702 # we couldn't find the end of this imageLink, so output it raw
1703 #but don't ignore what might be perfectly normal links in the text we've examined
1704 $text = $this->replaceInternalLinks($text);
1705 $s .= "{$prefix}[[$link|$text";
1706 # note: no $trail, because without an end, there *is* no trail
1707 wfProfileOut( "$fname-might_be_img" );
1710 } else { #it's not an image, so output it raw
1711 $s .= "{$prefix}[[$link|$text";
1712 # note: no $trail, because without an end, there *is* no trail
1713 wfProfileOut( "$fname-might_be_img" );
1716 wfProfileOut( "$fname-might_be_img" );
1719 $wasblank = ( '' == $text );
1720 if( $wasblank ) $text = $link;
1722 # Link not escaped by : , create the various objects
1726 wfProfileIn( "$fname-interwiki" );
1727 if( $iw && $this->mOptions
->getInterwikiMagic() && $nottalk && $wgContLang->getLanguageName( $iw ) ) {
1728 $this->mOutput
->addLanguageLink( $nt->getFullText() );
1729 $s = rtrim($s . $prefix);
1730 $s .= trim($trail, "\n") == '' ?
'': $prefix . $trail;
1731 wfProfileOut( "$fname-interwiki" );
1734 wfProfileOut( "$fname-interwiki" );
1736 if ( $ns == NS_IMAGE
) {
1737 wfProfileIn( "$fname-image" );
1738 if ( !wfIsBadImage( $nt->getDBkey(), $this->mTitle
) ) {
1739 # recursively parse links inside the image caption
1740 # actually, this will parse them in any other parameters, too,
1741 # but it might be hard to fix that, and it doesn't matter ATM
1742 $text = $this->replaceExternalLinks($text);
1743 $text = $this->replaceInternalLinks($text);
1745 # cloak any absolute URLs inside the image markup, so replaceExternalLinks() won't touch them
1746 $s .= $prefix . $this->armorLinks( $this->makeImage( $nt, $text ) ) . $trail;
1747 $this->mOutput
->addImage( $nt->getDBkey() );
1749 wfProfileOut( "$fname-image" );
1752 # We still need to record the image's presence on the page
1753 $this->mOutput
->addImage( $nt->getDBkey() );
1755 wfProfileOut( "$fname-image" );
1759 if ( $ns == NS_CATEGORY
) {
1760 wfProfileIn( "$fname-category" );
1761 $s = rtrim($s . "\n"); # bug 87
1764 $sortkey = $this->getDefaultSort();
1768 $sortkey = Sanitizer
::decodeCharReferences( $sortkey );
1769 $sortkey = str_replace( "\n", '', $sortkey );
1770 $sortkey = $wgContLang->convertCategoryKey( $sortkey );
1771 $this->mOutput
->addCategory( $nt->getDBkey(), $sortkey );
1774 * Strip the whitespace Category links produce, see bug 87
1775 * @todo We might want to use trim($tmp, "\n") here.
1777 $s .= trim($prefix . $trail, "\n") == '' ?
'': $prefix . $trail;
1779 wfProfileOut( "$fname-category" );
1784 # Self-link checking
1785 if( $nt->getFragment() === '' ) {
1786 if( in_array( $nt->getPrefixedText(), $selflink, true ) ) {
1787 $s .= $prefix . $sk->makeSelfLinkObj( $nt, $text, '', $trail );
1792 # Special and Media are pseudo-namespaces; no pages actually exist in them
1793 if( $ns == NS_MEDIA
) {
1794 $link = $sk->makeMediaLinkObj( $nt, $text );
1795 # Cloak with NOPARSE to avoid replacement in replaceExternalLinks
1796 $s .= $prefix . $this->armorLinks( $link ) . $trail;
1797 $this->mOutput
->addImage( $nt->getDBkey() );
1799 } elseif( $ns == NS_SPECIAL
) {
1800 $s .= $this->makeKnownLinkHolder( $nt, $text, '', $trail, $prefix );
1802 } elseif( $ns == NS_IMAGE
) {
1803 $img = new Image( $nt );
1804 if( $img->exists() ) {
1805 // Force a blue link if the file exists; may be a remote
1806 // upload on the shared repository, and we want to see its
1807 // auto-generated page.
1808 $s .= $this->makeKnownLinkHolder( $nt, $text, '', $trail, $prefix );
1809 $this->mOutput
->addLink( $nt );
1813 $s .= $this->makeLinkHolder( $nt, $text, '', $trail, $prefix );
1815 wfProfileOut( $fname );
1820 * Make a link placeholder. The text returned can be later resolved to a real link with
1821 * replaceLinkHolders(). This is done for two reasons: firstly to avoid further
1822 * parsing of interwiki links, and secondly to allow all existence checks and
1823 * article length checks (for stub links) to be bundled into a single query.
1826 function makeLinkHolder( &$nt, $text = '', $query = '', $trail = '', $prefix = '' ) {
1827 wfProfileIn( __METHOD__
);
1828 if ( ! is_object($nt) ) {
1830 $retVal = "<!-- ERROR -->{$prefix}{$text}{$trail}";
1832 # Separate the link trail from the rest of the link
1833 list( $inside, $trail ) = Linker
::splitTrail( $trail );
1835 if ( $nt->isExternal() ) {
1836 $nr = array_push( $this->mInterwikiLinkHolders
['texts'], $prefix.$text.$inside );
1837 $this->mInterwikiLinkHolders
['titles'][] = $nt;
1838 $retVal = '<!--IWLINK '. ($nr-1) ."-->{$trail}";
1840 $nr = array_push( $this->mLinkHolders
['namespaces'], $nt->getNamespace() );
1841 $this->mLinkHolders
['dbkeys'][] = $nt->getDBkey();
1842 $this->mLinkHolders
['queries'][] = $query;
1843 $this->mLinkHolders
['texts'][] = $prefix.$text.$inside;
1844 $this->mLinkHolders
['titles'][] = $nt;
1846 $retVal = '<!--LINK '. ($nr-1) ."-->{$trail}";
1849 wfProfileOut( __METHOD__
);
1854 * Render a forced-blue link inline; protect against double expansion of
1855 * URLs if we're in a mode that prepends full URL prefixes to internal links.
1856 * Since this little disaster has to split off the trail text to avoid
1857 * breaking URLs in the following text without breaking trails on the
1858 * wiki links, it's been made into a horrible function.
1861 * @param string $text
1862 * @param string $query
1863 * @param string $trail
1864 * @param string $prefix
1865 * @return string HTML-wikitext mix oh yuck
1867 function makeKnownLinkHolder( $nt, $text = '', $query = '', $trail = '', $prefix = '' ) {
1868 list( $inside, $trail ) = Linker
::splitTrail( $trail );
1869 $sk = $this->mOptions
->getSkin();
1870 $link = $sk->makeKnownLinkObj( $nt, $text, $query, $inside, $prefix );
1871 return $this->armorLinks( $link ) . $trail;
1875 * Insert a NOPARSE hacky thing into any inline links in a chunk that's
1876 * going to go through further parsing steps before inline URL expansion.
1878 * In particular this is important when using action=render, which causes
1879 * full URLs to be included.
1881 * Oh man I hate our multi-layer parser!
1883 * @param string more-or-less HTML
1884 * @return string less-or-more HTML with NOPARSE bits
1886 function armorLinks( $text ) {
1887 return preg_replace( '/\b(' . wfUrlProtocols() . ')/',
1888 "{$this->mUniqPrefix}NOPARSE$1", $text );
1892 * Return true if subpage links should be expanded on this page.
1895 function areSubpagesAllowed() {
1896 # Some namespaces don't allow subpages
1897 global $wgNamespacesWithSubpages;
1898 return !empty($wgNamespacesWithSubpages[$this->mTitle
->getNamespace()]);
1902 * Handle link to subpage if necessary
1903 * @param string $target the source of the link
1904 * @param string &$text the link text, modified as necessary
1905 * @return string the full name of the link
1908 function maybeDoSubpageLink($target, &$text) {
1911 # :Foobar -- override special treatment of prefix (images, language links)
1912 # /Foobar -- convert to CurrentPage/Foobar
1913 # /Foobar/ -- convert to CurrentPage/Foobar, strip the initial / from text
1914 # ../ -- convert to CurrentPage, from CurrentPage/CurrentSubPage
1915 # ../Foobar -- convert to CurrentPage/Foobar, from CurrentPage/CurrentSubPage
1917 $fname = 'Parser::maybeDoSubpageLink';
1918 wfProfileIn( $fname );
1919 $ret = $target; # default return value is no change
1922 $target = trim( $target );
1924 # Some namespaces don't allow subpages,
1925 # so only perform processing if subpages are allowed
1926 if( $this->areSubpagesAllowed() ) {
1927 # Look at the first character
1928 if( $target != '' && $target{0} == '/' ) {
1929 # / at end means we don't want the slash to be shown
1930 $trailingSlashes = preg_match_all( '%(/+)$%', $target, $m );
1931 if( $trailingSlashes ) {
1932 $noslash = $target = substr( $target, 1, -strlen($m[0][0]) );
1934 $noslash = substr( $target, 1 );
1937 $ret = $this->mTitle
->getPrefixedText(). '/' . trim($noslash);
1938 if( '' === $text ) {
1940 } # this might be changed for ugliness reasons
1942 # check for .. subpage backlinks
1944 $nodotdot = $target;
1945 while( strncmp( $nodotdot, "../", 3 ) == 0 ) {
1947 $nodotdot = substr( $nodotdot, 3 );
1949 if($dotdotcount > 0) {
1950 $exploded = explode( '/', $this->mTitle
->GetPrefixedText() );
1951 if( count( $exploded ) > $dotdotcount ) { # not allowed to go below top level page
1952 $ret = implode( '/', array_slice( $exploded, 0, -$dotdotcount ) );
1953 # / at the end means don't show full path
1954 if( substr( $nodotdot, -1, 1 ) == '/' ) {
1955 $nodotdot = substr( $nodotdot, 0, -1 );
1956 if( '' === $text ) {
1960 $nodotdot = trim( $nodotdot );
1961 if( $nodotdot != '' ) {
1962 $ret .= '/' . $nodotdot;
1969 wfProfileOut( $fname );
1974 * Used by doBlockLevels()
1977 /* private */ function closeParagraph() {
1979 if ( '' != $this->mLastSection
) {
1980 $result = '</' . $this->mLastSection
. ">\n";
1982 $this->mInPre
= false;
1983 $this->mLastSection
= '';
1986 # getCommon() returns the length of the longest common substring
1987 # of both arguments, starting at the beginning of both.
1989 /* private */ function getCommon( $st1, $st2 ) {
1990 $fl = strlen( $st1 );
1991 $shorter = strlen( $st2 );
1992 if ( $fl < $shorter ) { $shorter = $fl; }
1994 for ( $i = 0; $i < $shorter; ++
$i ) {
1995 if ( $st1{$i} != $st2{$i} ) { break; }
1999 # These next three functions open, continue, and close the list
2000 # element appropriate to the prefix character passed into them.
2002 /* private */ function openList( $char ) {
2003 $result = $this->closeParagraph();
2005 if ( '*' == $char ) { $result .= '<ul><li>'; }
2006 else if ( '#' == $char ) { $result .= '<ol><li>'; }
2007 else if ( ':' == $char ) { $result .= '<dl><dd>'; }
2008 else if ( ';' == $char ) {
2009 $result .= '<dl><dt>';
2010 $this->mDTopen
= true;
2012 else { $result = '<!-- ERR 1 -->'; }
2017 /* private */ function nextItem( $char ) {
2018 if ( '*' == $char ||
'#' == $char ) { return '</li><li>'; }
2019 else if ( ':' == $char ||
';' == $char ) {
2021 if ( $this->mDTopen
) { $close = '</dt>'; }
2022 if ( ';' == $char ) {
2023 $this->mDTopen
= true;
2024 return $close . '<dt>';
2026 $this->mDTopen
= false;
2027 return $close . '<dd>';
2030 return '<!-- ERR 2 -->';
2033 /* private */ function closeList( $char ) {
2034 if ( '*' == $char ) { $text = '</li></ul>'; }
2035 else if ( '#' == $char ) { $text = '</li></ol>'; }
2036 else if ( ':' == $char ) {
2037 if ( $this->mDTopen
) {
2038 $this->mDTopen
= false;
2039 $text = '</dt></dl>';
2041 $text = '</dd></dl>';
2044 else { return '<!-- ERR 3 -->'; }
2050 * Make lists from lines starting with ':', '*', '#', etc.
2053 * @return string the lists rendered as HTML
2055 function doBlockLevels( $text, $linestart ) {
2056 $fname = 'Parser::doBlockLevels';
2057 wfProfileIn( $fname );
2059 # Parsing through the text line by line. The main thing
2060 # happening here is handling of block-level elements p, pre,
2061 # and making lists from lines starting with * # : etc.
2063 $textLines = explode( "\n", $text );
2065 $lastPrefix = $output = '';
2066 $this->mDTopen
= $inBlockElem = false;
2068 $paragraphStack = false;
2070 if ( !$linestart ) {
2071 $output .= array_shift( $textLines );
2073 foreach ( $textLines as $oLine ) {
2074 $lastPrefixLength = strlen( $lastPrefix );
2075 $preCloseMatch = preg_match('/<\\/pre/i', $oLine );
2076 $preOpenMatch = preg_match('/<pre/i', $oLine );
2077 if ( !$this->mInPre
) {
2078 # Multiple prefixes may abut each other for nested lists.
2079 $prefixLength = strspn( $oLine, '*#:;' );
2080 $pref = substr( $oLine, 0, $prefixLength );
2083 $pref2 = str_replace( ';', ':', $pref );
2084 $t = substr( $oLine, $prefixLength );
2085 $this->mInPre
= !empty($preOpenMatch);
2087 # Don't interpret any other prefixes in preformatted text
2089 $pref = $pref2 = '';
2094 if( $prefixLength && 0 == strcmp( $lastPrefix, $pref2 ) ) {
2095 # Same as the last item, so no need to deal with nesting or opening stuff
2096 $output .= $this->nextItem( substr( $pref, -1 ) );
2097 $paragraphStack = false;
2099 if ( substr( $pref, -1 ) == ';') {
2100 # The one nasty exception: definition lists work like this:
2101 # ; title : definition text
2102 # So we check for : in the remainder text to split up the
2103 # title and definition, without b0rking links.
2105 if ($this->findColonNoLinks($t, $term, $t2) !== false) {
2107 $output .= $term . $this->nextItem( ':' );
2110 } elseif( $prefixLength ||
$lastPrefixLength ) {
2111 # Either open or close a level...
2112 $commonPrefixLength = $this->getCommon( $pref, $lastPrefix );
2113 $paragraphStack = false;
2115 while( $commonPrefixLength < $lastPrefixLength ) {
2116 $output .= $this->closeList( $lastPrefix{$lastPrefixLength-1} );
2117 --$lastPrefixLength;
2119 if ( $prefixLength <= $commonPrefixLength && $commonPrefixLength > 0 ) {
2120 $output .= $this->nextItem( $pref{$commonPrefixLength-1} );
2122 while ( $prefixLength > $commonPrefixLength ) {
2123 $char = substr( $pref, $commonPrefixLength, 1 );
2124 $output .= $this->openList( $char );
2126 if ( ';' == $char ) {
2127 # FIXME: This is dupe of code above
2128 if ($this->findColonNoLinks($t, $term, $t2) !== false) {
2130 $output .= $term . $this->nextItem( ':' );
2133 ++
$commonPrefixLength;
2135 $lastPrefix = $pref2;
2137 if( 0 == $prefixLength ) {
2138 wfProfileIn( "$fname-paragraph" );
2139 # No prefix (not in list)--go to paragraph mode
2140 // XXX: use a stack for nestable elements like span, table and div
2141 $openmatch = preg_match('/(?:<table|<blockquote|<h1|<h2|<h3|<h4|<h5|<h6|<pre|<tr|<p|<ul|<ol|<li|<\\/tr|<\\/td|<\\/th)/iS', $t );
2142 $closematch = preg_match(
2143 '/(?:<\\/table|<\\/blockquote|<\\/h1|<\\/h2|<\\/h3|<\\/h4|<\\/h5|<\\/h6|'.
2144 '<td|<th|<\\/?div|<hr|<\\/pre|<\\/p|'.$this->mUniqPrefix
.'-pre|<\\/li|<\\/ul|<\\/ol|<\\/?center)/iS', $t );
2145 if ( $openmatch or $closematch ) {
2146 $paragraphStack = false;
2147 #Â TODO bug 5718: paragraph closed
2148 $output .= $this->closeParagraph();
2149 if ( $preOpenMatch and !$preCloseMatch ) {
2150 $this->mInPre
= true;
2152 if ( $closematch ) {
2153 $inBlockElem = false;
2155 $inBlockElem = true;
2157 } else if ( !$inBlockElem && !$this->mInPre
) {
2158 if ( ' ' == $t{0} and ( $this->mLastSection
== 'pre' or trim($t) != '' ) ) {
2160 if ($this->mLastSection
!= 'pre') {
2161 $paragraphStack = false;
2162 $output .= $this->closeParagraph().'<pre>';
2163 $this->mLastSection
= 'pre';
2165 $t = substr( $t, 1 );
2168 if ( '' == trim($t) ) {
2169 if ( $paragraphStack ) {
2170 $output .= $paragraphStack.'<br />';
2171 $paragraphStack = false;
2172 $this->mLastSection
= 'p';
2174 if ($this->mLastSection
!= 'p' ) {
2175 $output .= $this->closeParagraph();
2176 $this->mLastSection
= '';
2177 $paragraphStack = '<p>';
2179 $paragraphStack = '</p><p>';
2183 if ( $paragraphStack ) {
2184 $output .= $paragraphStack;
2185 $paragraphStack = false;
2186 $this->mLastSection
= 'p';
2187 } else if ($this->mLastSection
!= 'p') {
2188 $output .= $this->closeParagraph().'<p>';
2189 $this->mLastSection
= 'p';
2194 wfProfileOut( "$fname-paragraph" );
2196 // somewhere above we forget to get out of pre block (bug 785)
2197 if($preCloseMatch && $this->mInPre
) {
2198 $this->mInPre
= false;
2200 if ($paragraphStack === false) {
2204 while ( $prefixLength ) {
2205 $output .= $this->closeList( $pref2{$prefixLength-1} );
2208 if ( '' != $this->mLastSection
) {
2209 $output .= '</' . $this->mLastSection
. '>';
2210 $this->mLastSection
= '';
2213 wfProfileOut( $fname );
2218 * Split up a string on ':', ignoring any occurences inside tags
2219 * to prevent illegal overlapping.
2220 * @param string $str the string to split
2221 * @param string &$before set to everything before the ':'
2222 * @param string &$after set to everything after the ':'
2223 * return string the position of the ':', or false if none found
2225 function findColonNoLinks($str, &$before, &$after) {
2226 $fname = 'Parser::findColonNoLinks';
2227 wfProfileIn( $fname );
2229 $pos = strpos( $str, ':' );
2230 if( $pos === false ) {
2232 wfProfileOut( $fname );
2236 $lt = strpos( $str, '<' );
2237 if( $lt === false ||
$lt > $pos ) {
2238 // Easy; no tag nesting to worry about
2239 $before = substr( $str, 0, $pos );
2240 $after = substr( $str, $pos+
1 );
2241 wfProfileOut( $fname );
2245 // Ugly state machine to walk through avoiding tags.
2246 $state = MW_COLON_STATE_TEXT
;
2248 $len = strlen( $str );
2249 for( $i = 0; $i < $len; $i++
) {
2253 // (Using the number is a performance hack for common cases)
2254 case 0: // MW_COLON_STATE_TEXT:
2257 // Could be either a <start> tag or an </end> tag
2258 $state = MW_COLON_STATE_TAGSTART
;
2263 $before = substr( $str, 0, $i );
2264 $after = substr( $str, $i +
1 );
2265 wfProfileOut( $fname );
2268 // Embedded in a tag; don't break it.
2271 // Skip ahead looking for something interesting
2272 $colon = strpos( $str, ':', $i );
2273 if( $colon === false ) {
2274 // Nothing else interesting
2275 wfProfileOut( $fname );
2278 $lt = strpos( $str, '<', $i );
2279 if( $stack === 0 ) {
2280 if( $lt === false ||
$colon < $lt ) {
2282 $before = substr( $str, 0, $colon );
2283 $after = substr( $str, $colon +
1 );
2284 wfProfileOut( $fname );
2288 if( $lt === false ) {
2289 // Nothing else interesting to find; abort!
2290 // We're nested, but there's no close tags left. Abort!
2293 // Skip ahead to next tag start
2295 $state = MW_COLON_STATE_TAGSTART
;
2298 case 1: // MW_COLON_STATE_TAG:
2303 $state = MW_COLON_STATE_TEXT
;
2306 // Slash may be followed by >?
2307 $state = MW_COLON_STATE_TAGSLASH
;
2313 case 2: // MW_COLON_STATE_TAGSTART:
2316 $state = MW_COLON_STATE_CLOSETAG
;
2319 $state = MW_COLON_STATE_COMMENT
;
2322 // Illegal early close? This shouldn't happen D:
2323 $state = MW_COLON_STATE_TEXT
;
2326 $state = MW_COLON_STATE_TAG
;
2329 case 3: // MW_COLON_STATE_CLOSETAG:
2334 wfDebug( "Invalid input in $fname; too many close tags\n" );
2335 wfProfileOut( $fname );
2338 $state = MW_COLON_STATE_TEXT
;
2341 case MW_COLON_STATE_TAGSLASH
:
2343 // Yes, a self-closed tag <blah/>
2344 $state = MW_COLON_STATE_TEXT
;
2346 // Probably we're jumping the gun, and this is an attribute
2347 $state = MW_COLON_STATE_TAG
;
2350 case 5: // MW_COLON_STATE_COMMENT:
2352 $state = MW_COLON_STATE_COMMENTDASH
;
2355 case MW_COLON_STATE_COMMENTDASH
:
2357 $state = MW_COLON_STATE_COMMENTDASHDASH
;
2359 $state = MW_COLON_STATE_COMMENT
;
2362 case MW_COLON_STATE_COMMENTDASHDASH
:
2364 $state = MW_COLON_STATE_TEXT
;
2366 $state = MW_COLON_STATE_COMMENT
;
2370 throw new MWException( "State machine error in $fname" );
2374 wfDebug( "Invalid input in $fname; not enough close tags (stack $stack, state $state)\n" );
2377 wfProfileOut( $fname );
2382 * Return value of a magic variable (like PAGENAME)
2386 function getVariableValue( $index ) {
2387 global $wgContLang, $wgSitename, $wgServer, $wgServerName, $wgScriptPath;
2390 * Some of these require message or data lookups and can be
2391 * expensive to check many times.
2393 static $varCache = array();
2394 if ( wfRunHooks( 'ParserGetVariableValueVarCache', array( &$this, &$varCache ) ) ) {
2395 if ( isset( $varCache[$index] ) ) {
2396 return $varCache[$index];
2401 wfRunHooks( 'ParserGetVariableValueTs', array( &$this, &$ts ) );
2404 global $wgLocaltimezone;
2405 if ( isset( $wgLocaltimezone ) ) {
2406 $oldtz = getenv( 'TZ' );
2407 putenv( 'TZ='.$wgLocaltimezone );
2409 $localTimestamp = date( 'YmdHis', $ts );
2410 $localMonth = date( 'm', $ts );
2411 $localMonthName = date( 'n', $ts );
2412 $localDay = date( 'j', $ts );
2413 $localDay2 = date( 'd', $ts );
2414 $localDayOfWeek = date( 'w', $ts );
2415 $localWeek = date( 'W', $ts );
2416 $localYear = date( 'Y', $ts );
2417 $localHour = date( 'H', $ts );
2418 if ( isset( $wgLocaltimezone ) ) {
2419 putenv( 'TZ='.$oldtz );
2423 case 'currentmonth':
2424 return $varCache[$index] = $wgContLang->formatNum( date( 'm', $ts ) );
2425 case 'currentmonthname':
2426 return $varCache[$index] = $wgContLang->getMonthName( date( 'n', $ts ) );
2427 case 'currentmonthnamegen':
2428 return $varCache[$index] = $wgContLang->getMonthNameGen( date( 'n', $ts ) );
2429 case 'currentmonthabbrev':
2430 return $varCache[$index] = $wgContLang->getMonthAbbreviation( date( 'n', $ts ) );
2432 return $varCache[$index] = $wgContLang->formatNum( date( 'j', $ts ) );
2434 return $varCache[$index] = $wgContLang->formatNum( date( 'd', $ts ) );
2436 return $varCache[$index] = $wgContLang->formatNum( $localMonth );
2437 case 'localmonthname':
2438 return $varCache[$index] = $wgContLang->getMonthName( $localMonthName );
2439 case 'localmonthnamegen':
2440 return $varCache[$index] = $wgContLang->getMonthNameGen( $localMonthName );
2441 case 'localmonthabbrev':
2442 return $varCache[$index] = $wgContLang->getMonthAbbreviation( $localMonthName );
2444 return $varCache[$index] = $wgContLang->formatNum( $localDay );
2446 return $varCache[$index] = $wgContLang->formatNum( $localDay2 );
2448 return $this->mTitle
->getText();
2450 return $this->mTitle
->getPartialURL();
2451 case 'fullpagename':
2452 return $this->mTitle
->getPrefixedText();
2453 case 'fullpagenamee':
2454 return $this->mTitle
->getPrefixedURL();
2456 return $this->mTitle
->getSubpageText();
2457 case 'subpagenamee':
2458 return $this->mTitle
->getSubpageUrlForm();
2459 case 'basepagename':
2460 return $this->mTitle
->getBaseText();
2461 case 'basepagenamee':
2462 return wfUrlEncode( str_replace( ' ', '_', $this->mTitle
->getBaseText() ) );
2463 case 'talkpagename':
2464 if( $this->mTitle
->canTalk() ) {
2465 $talkPage = $this->mTitle
->getTalkPage();
2466 return $talkPage->getPrefixedText();
2470 case 'talkpagenamee':
2471 if( $this->mTitle
->canTalk() ) {
2472 $talkPage = $this->mTitle
->getTalkPage();
2473 return $talkPage->getPrefixedUrl();
2477 case 'subjectpagename':
2478 $subjPage = $this->mTitle
->getSubjectPage();
2479 return $subjPage->getPrefixedText();
2480 case 'subjectpagenamee':
2481 $subjPage = $this->mTitle
->getSubjectPage();
2482 return $subjPage->getPrefixedUrl();
2484 return $this->mRevisionId
;
2486 return intval( substr( $this->getRevisionTimestamp(), 6, 2 ) );
2487 case 'revisionday2':
2488 return substr( $this->getRevisionTimestamp(), 6, 2 );
2489 case 'revisionmonth':
2490 return intval( substr( $this->getRevisionTimestamp(), 4, 2 ) );
2491 case 'revisionyear':
2492 return substr( $this->getRevisionTimestamp(), 0, 4 );
2493 case 'revisiontimestamp':
2494 return $this->getRevisionTimestamp();
2496 return str_replace('_',' ',$wgContLang->getNsText( $this->mTitle
->getNamespace() ) );
2498 return wfUrlencode( $wgContLang->getNsText( $this->mTitle
->getNamespace() ) );
2500 return $this->mTitle
->canTalk() ?
str_replace('_',' ',$this->mTitle
->getTalkNsText()) : '';
2502 return $this->mTitle
->canTalk() ?
wfUrlencode( $this->mTitle
->getTalkNsText() ) : '';
2503 case 'subjectspace':
2504 return $this->mTitle
->getSubjectNsText();
2505 case 'subjectspacee':
2506 return( wfUrlencode( $this->mTitle
->getSubjectNsText() ) );
2507 case 'currentdayname':
2508 return $varCache[$index] = $wgContLang->getWeekdayName( date( 'w', $ts ) +
1 );
2510 return $varCache[$index] = $wgContLang->formatNum( date( 'Y', $ts ), true );
2512 return $varCache[$index] = $wgContLang->time( wfTimestamp( TS_MW
, $ts ), false, false );
2514 return $varCache[$index] = $wgContLang->formatNum( date( 'H', $ts ), true );
2516 // @bug 4594 PHP5 has it zero padded, PHP4 does not, cast to
2517 // int to remove the padding
2518 return $varCache[$index] = $wgContLang->formatNum( (int)date( 'W', $ts ) );
2520 return $varCache[$index] = $wgContLang->formatNum( date( 'w', $ts ) );
2521 case 'localdayname':
2522 return $varCache[$index] = $wgContLang->getWeekdayName( $localDayOfWeek +
1 );
2524 return $varCache[$index] = $wgContLang->formatNum( $localYear, true );
2526 return $varCache[$index] = $wgContLang->time( $localTimestamp, false, false );
2528 return $varCache[$index] = $wgContLang->formatNum( $localHour, true );
2530 // @bug 4594 PHP5 has it zero padded, PHP4 does not, cast to
2531 // int to remove the padding
2532 return $varCache[$index] = $wgContLang->formatNum( (int)$localWeek );
2534 return $varCache[$index] = $wgContLang->formatNum( $localDayOfWeek );
2535 case 'numberofarticles':
2536 return $varCache[$index] = $wgContLang->formatNum( SiteStats
::articles() );
2537 case 'numberoffiles':
2538 return $varCache[$index] = $wgContLang->formatNum( SiteStats
::images() );
2539 case 'numberofusers':
2540 return $varCache[$index] = $wgContLang->formatNum( SiteStats
::users() );
2541 case 'numberofpages':
2542 return $varCache[$index] = $wgContLang->formatNum( SiteStats
::pages() );
2543 case 'numberofadmins':
2544 return $varCache[$index] = $wgContLang->formatNum( SiteStats
::admins() );
2545 case 'numberofedits':
2546 return $varCache[$index] = $wgContLang->formatNum( SiteStats
::edits() );
2547 case 'currenttimestamp':
2548 return $varCache[$index] = wfTimestampNow();
2549 case 'localtimestamp':
2550 return $varCache[$index] = $localTimestamp;
2551 case 'currentversion':
2552 return $varCache[$index] = SpecialVersion
::getVersion();
2558 return $wgServerName;
2560 return $wgScriptPath;
2561 case 'directionmark':
2562 return $wgContLang->getDirMark();
2563 case 'contentlanguage':
2564 global $wgContLanguageCode;
2565 return $wgContLanguageCode;
2568 if ( wfRunHooks( 'ParserGetVariableValueSwitch', array( &$this, &$varCache, &$index, &$ret ) ) )
2576 * initialise the magic variables (like CURRENTMONTHNAME)
2580 function initialiseVariables() {
2581 $fname = 'Parser::initialiseVariables';
2582 wfProfileIn( $fname );
2583 $variableIDs = MagicWord
::getVariableIDs();
2585 $this->mVariables
= array();
2586 foreach ( $variableIDs as $id ) {
2587 $mw =& MagicWord
::get( $id );
2588 $mw->addToArray( $this->mVariables
, $id );
2590 wfProfileOut( $fname );
2594 * parse any parentheses in format ((title|part|part))
2595 * and call callbacks to get a replacement text for any found piece
2597 * @param string $text The text to parse
2598 * @param array $callbacks rules in form:
2599 * '{' => array( # opening parentheses
2600 * 'end' => '}', # closing parentheses
2601 * 'cb' => array(2 => callback, # replacement callback to call if {{..}} is found
2602 * 3 => callback # replacement callback to call if {{{..}}} is found
2605 * 'min' => 2, # Minimum parenthesis count in cb
2606 * 'max' => 3, # Maximum parenthesis count in cb
2609 function replace_callback ($text, $callbacks) {
2610 wfProfileIn( __METHOD__
);
2611 $openingBraceStack = array(); # this array will hold a stack of parentheses which are not closed yet
2612 $lastOpeningBrace = -1; # last not closed parentheses
2614 $validOpeningBraces = implode( '', array_keys( $callbacks ) );
2617 while ( $i < strlen( $text ) ) {
2618 # Find next opening brace, closing brace or pipe
2619 if ( $lastOpeningBrace == -1 ) {
2620 $currentClosing = '';
2621 $search = $validOpeningBraces;
2623 $currentClosing = $openingBraceStack[$lastOpeningBrace]['braceEnd'];
2624 $search = $validOpeningBraces . '|' . $currentClosing;
2627 $i +
= strcspn( $text, $search, $i );
2628 if ( $i < strlen( $text ) ) {
2629 if ( $text[$i] == '|' ) {
2631 } elseif ( $text[$i] == $currentClosing ) {
2633 } elseif ( isset( $callbacks[$text[$i]] ) ) {
2635 $rule = $callbacks[$text[$i]];
2637 # Some versions of PHP have a strcspn which stops on null characters
2638 # Ignore and continue
2647 if ( $found == 'open' ) {
2648 # found opening brace, let's add it to parentheses stack
2649 $piece = array('brace' => $text[$i],
2650 'braceEnd' => $rule['end'],
2654 # count opening brace characters
2655 $piece['count'] = strspn( $text, $piece['brace'], $i );
2656 $piece['startAt'] = $piece['partStart'] = $i +
$piece['count'];
2657 $i +
= $piece['count'];
2659 # we need to add to stack only if opening brace count is enough for one of the rules
2660 if ( $piece['count'] >= $rule['min'] ) {
2661 $lastOpeningBrace ++
;
2662 $openingBraceStack[$lastOpeningBrace] = $piece;
2664 } elseif ( $found == 'close' ) {
2665 # lets check if it is enough characters for closing brace
2666 $maxCount = $openingBraceStack[$lastOpeningBrace]['count'];
2667 $count = strspn( $text, $text[$i], $i, $maxCount );
2669 # check for maximum matching characters (if there are 5 closing
2670 # characters, we will probably need only 3 - depending on the rules)
2672 $matchingCallback = null;
2673 $cbType = $callbacks[$openingBraceStack[$lastOpeningBrace]['brace']];
2674 if ( $count > $cbType['max'] ) {
2675 # The specified maximum exists in the callback array, unless the caller
2677 $matchingCount = $cbType['max'];
2679 # Count is less than the maximum
2680 # Skip any gaps in the callback array to find the true largest match
2681 # Need to use array_key_exists not isset because the callback can be null
2682 $matchingCount = $count;
2683 while ( $matchingCount > 0 && !array_key_exists( $matchingCount, $cbType['cb'] ) ) {
2688 if ($matchingCount <= 0) {
2692 $matchingCallback = $cbType['cb'][$matchingCount];
2694 # let's set a title or last part (if '|' was found)
2695 if (null === $openingBraceStack[$lastOpeningBrace]['parts']) {
2696 $openingBraceStack[$lastOpeningBrace]['title'] =
2697 substr($text, $openingBraceStack[$lastOpeningBrace]['partStart'],
2698 $i - $openingBraceStack[$lastOpeningBrace]['partStart']);
2700 $openingBraceStack[$lastOpeningBrace]['parts'][] =
2701 substr($text, $openingBraceStack[$lastOpeningBrace]['partStart'],
2702 $i - $openingBraceStack[$lastOpeningBrace]['partStart']);
2705 $pieceStart = $openingBraceStack[$lastOpeningBrace]['startAt'] - $matchingCount;
2706 $pieceEnd = $i +
$matchingCount;
2708 if( is_callable( $matchingCallback ) ) {
2710 'text' => substr($text, $pieceStart, $pieceEnd - $pieceStart),
2711 'title' => trim($openingBraceStack[$lastOpeningBrace]['title']),
2712 'parts' => $openingBraceStack[$lastOpeningBrace]['parts'],
2713 'lineStart' => (($pieceStart > 0) && ($text[$pieceStart-1] == "\n")),
2715 # finally we can call a user callback and replace piece of text
2716 $replaceWith = call_user_func( $matchingCallback, $cbArgs );
2717 $text = substr($text, 0, $pieceStart) . $replaceWith . substr($text, $pieceEnd);
2718 $i = $pieceStart +
strlen($replaceWith);
2720 # null value for callback means that parentheses should be parsed, but not replaced
2721 $i +
= $matchingCount;
2724 # reset last opening parentheses, but keep it in case there are unused characters
2725 $piece = array('brace' => $openingBraceStack[$lastOpeningBrace]['brace'],
2726 'braceEnd' => $openingBraceStack[$lastOpeningBrace]['braceEnd'],
2727 'count' => $openingBraceStack[$lastOpeningBrace]['count'],
2730 'startAt' => $openingBraceStack[$lastOpeningBrace]['startAt']);
2731 $openingBraceStack[$lastOpeningBrace--] = null;
2733 if ($matchingCount < $piece['count']) {
2734 $piece['count'] -= $matchingCount;
2735 $piece['startAt'] -= $matchingCount;
2736 $piece['partStart'] = $piece['startAt'];
2737 # do we still qualify for any callback with remaining count?
2738 $currentCbList = $callbacks[$piece['brace']]['cb'];
2739 while ( $piece['count'] ) {
2740 if ( array_key_exists( $piece['count'], $currentCbList ) ) {
2741 $lastOpeningBrace++
;
2742 $openingBraceStack[$lastOpeningBrace] = $piece;
2748 } elseif ( $found == 'pipe' ) {
2749 # lets set a title if it is a first separator, or next part otherwise
2750 if (null === $openingBraceStack[$lastOpeningBrace]['parts']) {
2751 $openingBraceStack[$lastOpeningBrace]['title'] =
2752 substr($text, $openingBraceStack[$lastOpeningBrace]['partStart'],
2753 $i - $openingBraceStack[$lastOpeningBrace]['partStart']);
2754 $openingBraceStack[$lastOpeningBrace]['parts'] = array();
2756 $openingBraceStack[$lastOpeningBrace]['parts'][] =
2757 substr($text, $openingBraceStack[$lastOpeningBrace]['partStart'],
2758 $i - $openingBraceStack[$lastOpeningBrace]['partStart']);
2760 $openingBraceStack[$lastOpeningBrace]['partStart'] = ++
$i;
2764 wfProfileOut( __METHOD__
);
2769 * Replace magic variables, templates, and template arguments
2770 * with the appropriate text. Templates are substituted recursively,
2771 * taking care to avoid infinite loops.
2773 * Note that the substitution depends on value of $mOutputType:
2774 * OT_WIKI: only {{subst:}} templates
2775 * OT_MSG: only magic variables
2776 * OT_HTML: all templates and magic variables
2778 * @param string $tex The text to transform
2779 * @param array $args Key-value pairs representing template parameters to substitute
2780 * @param bool $argsOnly Only do argument (triple-brace) expansion, not double-brace expansion
2783 function replaceVariables( $text, $args = array(), $argsOnly = false ) {
2784 # Prevent too big inclusions
2785 if( strlen( $text ) > $this->mOptions
->getMaxIncludeSize() ) {
2789 $fname = __METHOD__
/*. '-L' . count( $this->mArgStack )*/;
2790 wfProfileIn( $fname );
2792 # This function is called recursively. To keep track of arguments we need a stack:
2793 array_push( $this->mArgStack
, $args );
2795 $braceCallbacks = array();
2797 $braceCallbacks[2] = array( &$this, 'braceSubstitution' );
2799 if ( $this->mOutputType
!= OT_MSG
) {
2800 $braceCallbacks[3] = array( &$this, 'argSubstitution' );
2802 if ( $braceCallbacks ) {
2806 'cb' => $braceCallbacks,
2807 'min' => $argsOnly ?
3 : 2,
2808 'max' => isset( $braceCallbacks[3] ) ?
3 : 2,
2812 'cb' => array(2=>null),
2817 $text = $this->replace_callback ($text, $callbacks);
2819 array_pop( $this->mArgStack
);
2821 wfProfileOut( $fname );
2826 * Replace magic variables
2829 function variableSubstitution( $matches ) {
2831 $fname = 'Parser::variableSubstitution';
2832 $varname = $wgContLang->lc($matches[1]);
2833 wfProfileIn( $fname );
2835 if ( $this->mOutputType
== OT_WIKI
) {
2836 # Do only magic variables prefixed by SUBST
2837 $mwSubst =& MagicWord
::get( 'subst' );
2838 if (!$mwSubst->matchStartAndRemove( $varname ))
2840 # Note that if we don't substitute the variable below,
2841 # we don't remove the {{subst:}} magic word, in case
2842 # it is a template rather than a magic variable.
2844 if ( !$skip && array_key_exists( $varname, $this->mVariables
) ) {
2845 $id = $this->mVariables
[$varname];
2846 # Now check if we did really match, case sensitive or not
2847 $mw =& MagicWord
::get( $id );
2848 if ($mw->match($matches[1])) {
2849 $text = $this->getVariableValue( $id );
2850 $this->mOutput
->mContainsOldMagic
= true;
2852 $text = $matches[0];
2855 $text = $matches[0];
2857 wfProfileOut( $fname );
2862 /// Clean up argument array - refactored in 1.9 so parserfunctions can use it, too.
2863 static function createAssocArgs( $args ) {
2864 $assocArgs = array();
2866 foreach( $args as $arg ) {
2867 $eqpos = strpos( $arg, '=' );
2868 if ( $eqpos === false ) {
2869 $assocArgs[$index++
] = $arg;
2871 $name = trim( substr( $arg, 0, $eqpos ) );
2872 $value = trim( substr( $arg, $eqpos+
1 ) );
2873 if ( $value === false ) {
2876 if ( $name !== false ) {
2877 $assocArgs[$name] = $value;
2886 * Return the text of a template, after recursively
2887 * replacing any variables or templates within the template.
2889 * @param array $piece The parts of the template
2890 * $piece['text']: matched text
2891 * $piece['title']: the title, i.e. the part before the |
2892 * $piece['parts']: the parameter array
2893 * @return string the text of the template
2896 function braceSubstitution( $piece ) {
2897 global $wgContLang, $wgLang, $wgAllowDisplayTitle, $wgNonincludableNamespaces;
2898 $fname = __METHOD__
/*. '-L' . count( $this->mArgStack )*/;
2899 wfProfileIn( $fname );
2900 wfProfileIn( __METHOD__
.'-setup' );
2903 $found = false; # $text has been filled
2904 $nowiki = false; # wiki markup in $text should be escaped
2905 $noparse = false; # Unsafe HTML tags should not be stripped, etc.
2906 $noargs = false; # Don't replace triple-brace arguments in $text
2907 $replaceHeadings = false; # Make the edit section links go to the template not the article
2908 $headingOffset = 0; # Skip headings when number, to account for those that weren't transcluded.
2909 $isHTML = false; # $text is HTML, armour it against wikitext transformation
2910 $forceRawInterwiki = false; # Force interwiki transclusion to be done in raw mode not rendered
2912 # Title object, where $text came from
2918 # $part1 is the bit before the first |, and must contain only title characters
2919 # $args is a list of arguments, starting from index 0, not including $part1
2921 $titleText = $part1 = $piece['title'];
2922 # If the third subpattern matched anything, it will start with |
2924 if (null == $piece['parts']) {
2925 $replaceWith = $this->variableSubstitution (array ($piece['text'], $piece['title']));
2926 if ($replaceWith != $piece['text']) {
2927 $text = $replaceWith;
2934 $args = (null == $piece['parts']) ?
array() : $piece['parts'];
2935 wfProfileOut( __METHOD__
.'-setup' );
2938 wfProfileIn( __METHOD__
.'-modifiers' );
2940 $mwSubst =& MagicWord
::get( 'subst' );
2941 if ( $mwSubst->matchStartAndRemove( $part1 ) xor $this->ot
['wiki'] ) {
2942 # One of two possibilities is true:
2943 # 1) Found SUBST but not in the PST phase
2944 # 2) Didn't find SUBST and in the PST phase
2945 # In either case, return without further processing
2946 $text = $piece['text'];
2953 # MSG, MSGNW and RAW
2956 $mwMsgnw =& MagicWord
::get( 'msgnw' );
2957 if ( $mwMsgnw->matchStartAndRemove( $part1 ) ) {
2960 # Remove obsolete MSG:
2961 $mwMsg =& MagicWord
::get( 'msg' );
2962 $mwMsg->matchStartAndRemove( $part1 );
2966 $mwRaw =& MagicWord
::get( 'raw' );
2967 if ( $mwRaw->matchStartAndRemove( $part1 ) ) {
2968 $forceRawInterwiki = true;
2971 wfProfileOut( __METHOD__
.'-modifiers' );
2973 //save path level before recursing into functions & templates.
2974 $lastPathLevel = $this->mTemplatePath
;
2978 wfProfileIn( __METHOD__
. '-pfunc' );
2980 $colonPos = strpos( $part1, ':' );
2981 if ( $colonPos !== false ) {
2982 # Case sensitive functions
2983 $function = substr( $part1, 0, $colonPos );
2984 if ( isset( $this->mFunctionSynonyms
[1][$function] ) ) {
2985 $function = $this->mFunctionSynonyms
[1][$function];
2987 # Case insensitive functions
2988 $function = strtolower( $function );
2989 if ( isset( $this->mFunctionSynonyms
[0][$function] ) ) {
2990 $function = $this->mFunctionSynonyms
[0][$function];
2996 $funcArgs = array_map( 'trim', $args );
2997 $funcArgs = array_merge( array( &$this, trim( substr( $part1, $colonPos +
1 ) ) ), $funcArgs );
2998 $result = call_user_func_array( $this->mFunctionHooks
[$function], $funcArgs );
3001 // The text is usually already parsed, doesn't need triple-brace tags expanded, etc.
3005 if ( is_array( $result ) ) {
3006 if ( isset( $result[0] ) ) {
3007 $text = $linestart . $result[0];
3008 unset( $result[0] );
3011 // Extract flags into the local scope
3012 // This allows callers to set flags such as nowiki, noparse, found, etc.
3015 $text = $linestart . $result;
3019 wfProfileOut( __METHOD__
. '-pfunc' );
3022 # Template table test
3024 # Did we encounter this template already? If yes, it is in the cache
3025 # and we need to check for loops.
3026 if ( !$found && isset( $this->mTemplates
[$piece['title']] ) ) {
3029 # Infinite loop test
3030 if ( isset( $this->mTemplatePath
[$part1] ) ) {
3034 $text = $linestart .
3035 "[[$part1]]<!-- WARNING: template loop detected -->";
3036 wfDebug( __METHOD__
.": template loop broken at '$part1'\n" );
3038 # set $text to cached message.
3039 $text = $linestart . $this->mTemplates
[$piece['title']];
3040 #treat title for cached page the same as others
3043 $part1 = $this->maybeDoSubpageLink( $part1, $subpage );
3044 if ($subpage !== '') {
3045 $ns = $this->mTitle
->getNamespace();
3047 $title = Title
::newFromText( $part1, $ns );
3048 //used by include size checking
3049 $titleText = $title->getPrefixedText();
3050 //used by edit section links
3051 $replaceHeadings = true;
3056 # Load from database
3058 wfProfileIn( __METHOD__
. '-loadtpl' );
3060 # declaring $subpage directly in the function call
3061 # does not work correctly with references and breaks
3062 # {{/subpage}}-style inclusions
3064 $part1 = $this->maybeDoSubpageLink( $part1, $subpage );
3065 if ($subpage !== '') {
3066 $ns = $this->mTitle
->getNamespace();
3068 $title = Title
::newFromText( $part1, $ns );
3071 if ( !is_null( $title ) ) {
3072 $titleText = $title->getPrefixedText();
3073 # Check for language variants if the template is not found
3074 if($wgContLang->hasVariants() && $title->getArticleID() == 0){
3075 $wgContLang->findVariantLink($part1, $title);
3078 if ( !$title->isExternal() ) {
3079 if ( $title->getNamespace() == NS_SPECIAL
&& $this->mOptions
->getAllowSpecialInclusion() && $this->ot
['html'] ) {
3080 $text = SpecialPage
::capturePath( $title );
3081 if ( is_string( $text ) ) {
3086 $this->disableCache();
3088 } else if ( $wgNonincludableNamespaces && in_array( $title->getNamespace(), $wgNonincludableNamespaces ) ) {
3089 $found = false; //access denied
3090 wfDebug( "$fname: template inclusion denied for " . $title->getPrefixedDBkey() );
3092 $articleContent = $this->fetchTemplate( $title );
3093 if ( $articleContent !== false ) {
3095 $text = $articleContent;
3096 $replaceHeadings = true;
3100 # If the title is valid but undisplayable, make a link to it
3101 if ( !$found && ( $this->ot
['html'] ||
$this->ot
['pre'] ) ) {
3102 $text = "[[:$titleText]]";
3105 } elseif ( $title->isTrans() ) {
3106 // Interwiki transclusion
3107 if ( $this->ot
['html'] && !$forceRawInterwiki ) {
3108 $text = $this->interwikiTransclude( $title, 'render' );
3112 $text = $this->interwikiTransclude( $title, 'raw' );
3113 $replaceHeadings = true;
3118 # Template cache array insertion
3119 # Use the original $piece['title'] not the mangled $part1, so that
3120 # modifiers such as RAW: produce separate cache entries
3123 // A special page; don't store it in the template cache.
3125 $this->mTemplates
[$piece['title']] = $text;
3127 $text = $linestart . $text;
3130 wfProfileOut( __METHOD__
. '-loadtpl' );
3133 if ( $found && !$this->incrementIncludeSize( 'pre-expand', strlen( $text ) ) ) {
3134 # Error, oversize inclusion
3135 $text = $linestart .
3136 "[[$titleText]]<!-- WARNING: template omitted, pre-expand include size too large -->";
3141 # Recursive parsing, escaping and link table handling
3142 # Only for HTML output
3143 if ( $nowiki && $found && ( $this->ot
['html'] ||
$this->ot
['pre'] ) ) {
3144 $text = wfEscapeWikiText( $text );
3145 } elseif ( !$this->ot
['msg'] && $found ) {
3147 $assocArgs = array();
3149 # Clean up argument array
3150 $assocArgs = self
::createAssocArgs($args);
3151 # Add a new element to the templace recursion path
3152 $this->mTemplatePath
[$part1] = 1;
3156 # If there are any <onlyinclude> tags, only include them
3157 if ( in_string( '<onlyinclude>', $text ) && in_string( '</onlyinclude>', $text ) ) {
3158 $replacer = new OnlyIncludeReplacer
;
3159 StringUtils
::delimiterReplaceCallback( '<onlyinclude>', '</onlyinclude>',
3160 array( &$replacer, 'replace' ), $text );
3161 $text = $replacer->output
;
3163 # Remove <noinclude> sections and <includeonly> tags
3164 $text = StringUtils
::delimiterReplace( '<noinclude>', '</noinclude>', '', $text );
3165 $text = strtr( $text, array( '<includeonly>' => '' , '</includeonly>' => '' ) );
3167 if( $this->ot
['html'] ||
$this->ot
['pre'] ) {
3168 # Strip <nowiki>, <pre>, etc.
3169 $text = $this->strip( $text, $this->mStripState
);
3170 if ( $this->ot
['html'] ) {
3171 $text = Sanitizer
::removeHTMLtags( $text, array( &$this, 'replaceVariables' ), $assocArgs );
3172 } elseif ( $this->ot
['pre'] && $this->mOptions
->getRemoveComments() ) {
3173 $text = Sanitizer
::removeHTMLcomments( $text );
3176 $text = $this->replaceVariables( $text, $assocArgs );
3178 # If the template begins with a table or block-level
3179 # element, it should be treated as beginning a new line.
3180 if (!$piece['lineStart'] && preg_match('/^(?:{\\||:|;|#|\*)/', $text)) /*}*/{
3181 $text = "\n" . $text;
3183 } elseif ( !$noargs ) {
3184 # $noparse and !$noargs
3185 # Just replace the arguments, not any double-brace items
3186 # This is used for rendered interwiki transclusion
3187 $text = $this->replaceVariables( $text, $assocArgs, true );
3190 # Prune lower levels off the recursion check path
3191 $this->mTemplatePath
= $lastPathLevel;
3193 if ( $found && !$this->incrementIncludeSize( 'post-expand', strlen( $text ) ) ) {
3194 # Error, oversize inclusion
3195 $text = $linestart .
3196 "[[$titleText]]<!-- WARNING: template omitted, post-expand include size too large -->";
3202 wfProfileOut( $fname );
3203 return $piece['text'];
3205 wfProfileIn( __METHOD__
. '-placeholders' );
3207 # Replace raw HTML by a placeholder
3208 # Add a blank line preceding, to prevent it from mucking up
3209 # immediately preceding headings
3210 $text = "\n\n" . $this->insertStripItem( $text, $this->mStripState
);
3212 # replace ==section headers==
3213 # XXX this needs to go away once we have a better parser.
3214 if ( !$this->ot
['wiki'] && !$this->ot
['pre'] && $replaceHeadings ) {
3215 if( !is_null( $title ) )
3216 $encodedname = base64_encode($title->getPrefixedDBkey());
3218 $encodedname = base64_encode("");
3219 $m = preg_split('/(^={1,6}.*?={1,6}\s*?$)/m', $text, -1,
3220 PREG_SPLIT_DELIM_CAPTURE
);
3222 $nsec = $headingOffset;
3223 for( $i = 0; $i < count($m); $i +
= 2 ) {
3225 if (!isset($m[$i +
1]) ||
$m[$i +
1] == "") continue;
3227 if( strstr($hl, "<!--MWTEMPLATESECTION") ) {
3232 preg_match('/^(={1,6})(.*?)(={1,6})\s*?$/m', $hl, $m2);
3233 $text .= $m2[1] . $m2[2] . "<!--MWTEMPLATESECTION="
3234 . $encodedname . "&" . base64_encode("$nsec") . "-->" . $m2[3];
3240 wfProfileOut( __METHOD__
. '-placeholders' );
3243 # Prune lower levels off the recursion check path
3244 $this->mTemplatePath
= $lastPathLevel;
3247 wfProfileOut( $fname );
3248 return $piece['text'];
3250 wfProfileOut( $fname );
3256 * Fetch the unparsed text of a template and register a reference to it.
3258 function fetchTemplate( $title ) {
3260 // Loop to fetch the article, with up to 1 redirect
3261 for ( $i = 0; $i < 2 && is_object( $title ); $i++
) {
3262 $rev = Revision
::newFromTitle( $title );
3263 $this->mOutput
->addTemplate( $title, $title->getArticleID() );
3265 $text = $rev->getText();
3266 } elseif( $title->getNamespace() == NS_MEDIAWIKI
) {
3268 $message = $wgLang->lcfirst( $title->getText() );
3269 $text = wfMsgForContentNoTrans( $message );
3270 if( wfEmptyMsg( $message, $text ) ) {
3277 if ( $text === false ) {
3281 $title = Title
::newFromRedirect( $text );
3287 * Transclude an interwiki link.
3289 function interwikiTransclude( $title, $action ) {
3290 global $wgEnableScaryTranscluding;
3292 if (!$wgEnableScaryTranscluding)
3293 return wfMsg('scarytranscludedisabled');
3295 $url = $title->getFullUrl( "action=$action" );
3297 if (strlen($url) > 255)
3298 return wfMsg('scarytranscludetoolong');
3299 return $this->fetchScaryTemplateMaybeFromCache($url);
3302 function fetchScaryTemplateMaybeFromCache($url) {
3303 global $wgTranscludeCacheExpiry;
3304 $dbr = wfGetDB(DB_SLAVE
);
3305 $obj = $dbr->selectRow('transcache', array('tc_time', 'tc_contents'),
3306 array('tc_url' => $url));
3308 $time = $obj->tc_time
;
3309 $text = $obj->tc_contents
;
3310 if ($time && time() < $time +
$wgTranscludeCacheExpiry ) {
3315 $text = Http
::get($url);
3317 return wfMsg('scarytranscludefailed', $url);
3319 $dbw = wfGetDB(DB_MASTER
);
3320 $dbw->replace('transcache', array('tc_url'), array(
3322 'tc_time' => time(),
3323 'tc_contents' => $text));
3329 * Triple brace replacement -- used for template arguments
3332 function argSubstitution( $matches ) {
3333 $arg = trim( $matches['title'] );
3334 $text = $matches['text'];
3335 $inputArgs = end( $this->mArgStack
);
3337 if ( array_key_exists( $arg, $inputArgs ) ) {
3338 $text = $inputArgs[$arg];
3339 } else if (($this->mOutputType
== OT_HTML ||
$this->mOutputType
== OT_PREPROCESS
) &&
3340 null != $matches['parts'] && count($matches['parts']) > 0) {
3341 $text = $matches['parts'][0];
3343 if ( !$this->incrementIncludeSize( 'arg', strlen( $text ) ) ) {
3344 $text = $matches['text'] .
3345 '<!-- WARNING: argument omitted, expansion size too large -->';
3352 * Increment an include size counter
3354 * @param string $type The type of expansion
3355 * @param integer $size The size of the text
3356 * @return boolean False if this inclusion would take it over the maximum, true otherwise
3358 function incrementIncludeSize( $type, $size ) {
3359 if ( $this->mIncludeSizes
[$type] +
$size > $this->mOptions
->getMaxIncludeSize() ) {
3362 $this->mIncludeSizes
[$type] +
= $size;
3368 * Detect __NOGALLERY__ magic word and set a placeholder
3370 function stripNoGallery( &$text ) {
3371 # if the string __NOGALLERY__ (not case-sensitive) occurs in the HTML,
3373 $mw = MagicWord
::get( 'nogallery' );
3374 $this->mOutput
->mNoGallery
= $mw->matchAndRemove( $text ) ;
3378 * Detect __TOC__ magic word and set a placeholder
3380 function stripToc( $text ) {
3381 # if the string __NOTOC__ (not case-sensitive) occurs in the HTML,
3383 $mw = MagicWord
::get( 'notoc' );
3384 if( $mw->matchAndRemove( $text ) ) {
3385 $this->mShowToc
= false;
3388 $mw = MagicWord
::get( 'toc' );
3389 if( $mw->match( $text ) ) {
3390 $this->mShowToc
= true;
3391 $this->mForceTocPosition
= true;
3393 // Set a placeholder. At the end we'll fill it in with the TOC.
3394 $text = $mw->replace( '<!--MWTOC-->', $text, 1 );
3396 // Only keep the first one.
3397 $text = $mw->replace( '', $text );
3403 * This function accomplishes several tasks:
3404 * 1) Auto-number headings if that option is enabled
3405 * 2) Add an [edit] link to sections for logged in users who have enabled the option
3406 * 3) Add a Table of contents on the top for users who have enabled the option
3407 * 4) Auto-anchor headings
3409 * It loops through all headlines, collects the necessary data, then splits up the
3410 * string and re-inserts the newly formatted headlines.
3412 * @param string $text
3413 * @param boolean $isMain
3416 function formatHeadings( $text, $isMain=true ) {
3417 global $wgMaxTocLevel, $wgContLang;
3419 $doNumberHeadings = $this->mOptions
->getNumberHeadings();
3420 if( !$this->mTitle
->quickUserCan( 'edit' ) ) {
3423 $showEditLink = $this->mOptions
->getEditSection();
3426 # Inhibit editsection links if requested in the page
3427 $esw =& MagicWord
::get( 'noeditsection' );
3428 if( $esw->matchAndRemove( $text ) ) {
3432 # Get all headlines for numbering them and adding funky stuff like [edit]
3433 # links - this is for later, but we need the number of headlines right now
3435 $numMatches = preg_match_all( '/<H(?P<level>[1-6])(?P<attrib>.*?'.'>)(?P<header>.*?)<\/H[1-6] *>/i', $text, $matches );
3437 # if there are fewer than 4 headlines in the article, do not show TOC
3438 # unless it's been explicitly enabled.
3439 $enoughToc = $this->mShowToc
&&
3440 (($numMatches >= 4) ||
$this->mForceTocPosition
);
3442 # Allow user to stipulate that a page should have a "new section"
3443 # link added via __NEWSECTIONLINK__
3444 $mw =& MagicWord
::get( 'newsectionlink' );
3445 if( $mw->matchAndRemove( $text ) )
3446 $this->mOutput
->setNewSection( true );
3448 # if the string __FORCETOC__ (not case-sensitive) occurs in the HTML,
3449 # override above conditions and always show TOC above first header
3450 $mw =& MagicWord
::get( 'forcetoc' );
3451 if ($mw->matchAndRemove( $text ) ) {
3452 $this->mShowToc
= true;
3456 # We need this to perform operations on the HTML
3457 $sk = $this->mOptions
->getSkin();
3461 $sectionCount = 0; # headlineCount excluding template sections
3464 # Ugh .. the TOC should have neat indentation levels which can be
3465 # passed to the skin functions. These are determined here
3469 $sublevelCount = array();
3470 $levelCount = array();
3477 foreach( $matches[3] as $headline ) {
3479 $templatetitle = '';
3480 $templatesection = 0;
3483 if (preg_match("/<!--MWTEMPLATESECTION=([^&]+)&([^_]+)-->/", $headline, $mat)) {
3485 $templatetitle = base64_decode($mat[1]);
3486 $templatesection = 1 +
(int)base64_decode($mat[2]);
3487 $headline = preg_replace("/<!--MWTEMPLATESECTION=([^&]+)&([^_]+)-->/", "", $headline);
3491 $prevlevel = $level;
3493 $level = $matches[1][$headlineCount];
3495 if( $doNumberHeadings ||
$enoughToc ) {
3497 if ( $level > $prevlevel ) {
3498 # Increase TOC level
3500 $sublevelCount[$toclevel] = 0;
3501 if( $toclevel<$wgMaxTocLevel ) {
3502 $prevtoclevel = $toclevel;
3503 $toc .= $sk->tocIndent();
3507 elseif ( $level < $prevlevel && $toclevel > 1 ) {
3508 # Decrease TOC level, find level to jump to
3510 if ( $toclevel == 2 && $level <= $levelCount[1] ) {
3511 # Can only go down to level 1
3514 for ($i = $toclevel; $i > 0; $i--) {
3515 if ( $levelCount[$i] == $level ) {
3516 # Found last matching level
3520 elseif ( $levelCount[$i] < $level ) {
3521 # Found first matching level below current level
3527 if( $toclevel<$wgMaxTocLevel ) {
3528 $toc .= $sk->tocUnindent( $prevtoclevel - $toclevel );
3532 # No change in level, end TOC line
3533 if( $toclevel<$wgMaxTocLevel ) {
3534 $toc .= $sk->tocLineEnd();
3538 $levelCount[$toclevel] = $level;
3540 # count number of headlines for each level
3541 @$sublevelCount[$toclevel]++
;
3543 for( $i = 1; $i <= $toclevel; $i++
) {
3544 if( !empty( $sublevelCount[$i] ) ) {
3548 $numbering .= $wgContLang->formatNum( $sublevelCount[$i] );
3554 # The canonized header is a version of the header text safe to use for links
3555 # Avoid insertion of weird stuff like <math> by expanding the relevant sections
3556 $canonized_headline = $this->mStripState
->unstripBoth( $headline );
3558 # Remove link placeholders by the link text.
3559 # <!--LINK number-->
3561 # link text with suffix
3562 $canonized_headline = preg_replace( '/<!--LINK ([0-9]*)-->/e',
3563 "\$this->mLinkHolders['texts'][\$1]",
3564 $canonized_headline );
3565 $canonized_headline = preg_replace( '/<!--IWLINK ([0-9]*)-->/e',
3566 "\$this->mInterwikiLinkHolders['texts'][\$1]",
3567 $canonized_headline );
3570 $canonized_headline = preg_replace( '/<.*?' . '>/','',$canonized_headline );
3571 $tocline = trim( $canonized_headline );
3572 # Save headline for section edit hint before it's escaped
3573 $headline_hint = trim( $canonized_headline );
3574 $canonized_headline = Sanitizer
::escapeId( $tocline );
3575 $refers[$headlineCount] = $canonized_headline;
3577 # count how many in assoc. array so we can track dupes in anchors
3578 isset( $refers[$canonized_headline] ) ?
$refers[$canonized_headline]++
: $refers[$canonized_headline] = 1;
3579 $refcount[$headlineCount]=$refers[$canonized_headline];
3581 # Don't number the heading if it is the only one (looks silly)
3582 if( $doNumberHeadings && count( $matches[3] ) > 1) {
3583 # the two are different if the line contains a link
3584 $headline=$numbering . ' ' . $headline;
3587 # Create the anchor for linking from the TOC to the section
3588 $anchor = $canonized_headline;
3589 if($refcount[$headlineCount] > 1 ) {
3590 $anchor .= '_' . $refcount[$headlineCount];
3592 if( $enoughToc && ( !isset($wgMaxTocLevel) ||
$toclevel<$wgMaxTocLevel ) ) {
3593 $toc .= $sk->tocLine($anchor, $tocline, $numbering, $toclevel);
3595 # give headline the correct <h#> tag
3596 if( $showEditLink && ( !$istemplate ||
$templatetitle !== "" ) ) {
3598 $editlink = $sk->editSectionLinkForOther($templatetitle, $templatesection);
3600 $editlink = $sk->editSectionLink($this->mTitle
, $sectionCount+
1, $headline_hint);
3604 $head[$headlineCount] = $sk->makeHeadline( $level, $matches['attrib'][$headlineCount], $anchor, $headline, $editlink );
3611 # Never ever show TOC if no headers
3612 if( $numVisible < 1 ) {
3617 if( $prevtoclevel > 0 && $prevtoclevel < $wgMaxTocLevel ) {
3618 $toc .= $sk->tocUnindent( $prevtoclevel - 1 );
3620 $toc = $sk->tocList( $toc );
3623 # split up and insert constructed headlines
3625 $blocks = preg_split( '/<H[1-6].*?' . '>.*?<\/H[1-6]>/i', $text );
3628 foreach( $blocks as $block ) {
3629 if( $showEditLink && $headlineCount > 0 && $i == 0 && $block != "\n" ) {
3630 # This is the [edit] link that appears for the top block of text when
3631 # section editing is enabled
3633 # Disabled because it broke block formatting
3634 # For example, a bullet point in the top line
3635 # $full .= $sk->editSectionLink(0);
3638 if( $enoughToc && !$i && $isMain && !$this->mForceTocPosition
) {
3639 # Top anchor now in skin
3643 if( !empty( $head[$i] ) ) {
3648 if( $this->mForceTocPosition
) {
3649 return str_replace( '<!--MWTOC-->', $toc, $full );
3656 * Transform wiki markup when saving a page by doing \r\n -> \n
3657 * conversion, substitting signatures, {{subst:}} templates, etc.
3659 * @param string $text the text to transform
3660 * @param Title &$title the Title object for the current article
3661 * @param User &$user the User object describing the current user
3662 * @param ParserOptions $options parsing options
3663 * @param bool $clearState whether to clear the parser state first
3664 * @return string the altered wiki markup
3667 function preSaveTransform( $text, &$title, $user, $options, $clearState = true ) {
3668 $this->mOptions
= $options;
3669 $this->mTitle
=& $title;
3670 $this->setOutputType( OT_WIKI
);
3672 if ( $clearState ) {
3673 $this->clearState();
3676 $stripState = new StripState
;
3680 $text = str_replace( array_keys( $pairs ), array_values( $pairs ), $text );
3681 $text = $this->strip( $text, $stripState, true, array( 'gallery' ) );
3682 $text = $this->pstPass2( $text, $stripState, $user );
3683 $text = $stripState->unstripBoth( $text );
3688 * Pre-save transform helper function
3691 function pstPass2( $text, &$stripState, $user ) {
3692 global $wgContLang, $wgLocaltimezone;
3694 /* Note: This is the timestamp saved as hardcoded wikitext to
3695 * the database, we use $wgContLang here in order to give
3696 * everyone the same signature and use the default one rather
3697 * than the one selected in each user's preferences.
3699 if ( isset( $wgLocaltimezone ) ) {
3700 $oldtz = getenv( 'TZ' );
3701 putenv( 'TZ='.$wgLocaltimezone );
3703 $d = $wgContLang->timeanddate( date( 'YmdHis' ), false, false) .
3704 ' (' . date( 'T' ) . ')';
3705 if ( isset( $wgLocaltimezone ) ) {
3706 putenv( 'TZ='.$oldtz );
3709 # Variable replacement
3710 # Because mOutputType is OT_WIKI, this will only process {{subst:xxx}} type tags
3711 $text = $this->replaceVariables( $text );
3713 # Strip out <nowiki> etc. added via replaceVariables
3714 $text = $this->strip( $text, $stripState, false, array( 'gallery' ) );
3717 $sigText = $this->getUserSig( $user );
3718 $text = strtr( $text, array(
3720 '~~~~' => "$sigText $d",
3724 # Context links: [[|name]] and [[name (context)|]]
3726 global $wgLegalTitleChars;
3727 $tc = "[$wgLegalTitleChars]";
3728 $nc = '[ _0-9A-Za-z\x80-\xff]'; # Namespaces can use non-ascii!
3730 $p1 = "/\[\[(:?$nc+:|:|)($tc+?)( \\($tc+\\))\\|]]/"; # [[ns:page (context)|]]
3731 $p3 = "/\[\[(:?$nc+:|:|)($tc+?)( \\($tc+\\)|)(, $tc+|)\\|]]/"; # [[ns:page (context), context|]]
3732 $p2 = "/\[\[\\|($tc+)]]/"; # [[|page]]
3734 # try $p1 first, to turn "[[A, B (C)|]]" into "[[A, B (C)|A, B]]"
3735 $text = preg_replace( $p1, '[[\\1\\2\\3|\\2]]', $text );
3736 $text = preg_replace( $p3, '[[\\1\\2\\3\\4|\\2]]', $text );
3738 $t = $this->mTitle
->getText();
3740 if ( preg_match( "/^($nc+:|)$tc+?( \\($tc+\\))$/", $t, $m ) ) {
3741 $text = preg_replace( $p2, "[[$m[1]\\1$m[2]|\\1]]", $text );
3742 } elseif ( preg_match( "/^($nc+:|)$tc+?(, $tc+|)$/", $t, $m ) && '' != "$m[1]$m[2]" ) {
3743 $text = preg_replace( $p2, "[[$m[1]\\1$m[2]|\\1]]", $text );
3745 # if there's no context, don't bother duplicating the title
3746 $text = preg_replace( $p2, '[[\\1]]', $text );
3749 # Trim trailing whitespace
3750 $text = rtrim( $text );
3756 * Fetch the user's signature text, if any, and normalize to
3757 * validated, ready-to-insert wikitext.
3763 function getUserSig( &$user ) {
3764 $username = $user->getName();
3765 $nickname = $user->getOption( 'nickname' );
3766 $nickname = $nickname === '' ?
$username : $nickname;
3768 if( $user->getBoolOption( 'fancysig' ) !== false ) {
3769 # Sig. might contain markup; validate this
3770 if( $this->validateSig( $nickname ) !== false ) {
3771 # Validated; clean up (if needed) and return it
3772 return $this->cleanSig( $nickname, true );
3774 # Failed to validate; fall back to the default
3775 $nickname = $username;
3776 wfDebug( "Parser::getUserSig: $username has bad XML tags in signature.\n" );
3780 // Make sure nickname doesnt get a sig in a sig
3781 $nickname = $this->cleanSigInSig( $nickname );
3783 # If we're still here, make it a link to the user page
3784 $userpage = $user->getUserPage();
3785 return( '[[' . $userpage->getPrefixedText() . '|' . wfEscapeWikiText( $nickname ) . ']]' );
3789 * Check that the user's signature contains no bad XML
3791 * @param string $text
3792 * @return mixed An expanded string, or false if invalid.
3794 function validateSig( $text ) {
3795 return( wfIsWellFormedXmlFragment( $text ) ?
$text : false );
3799 * Clean up signature text
3801 * 1) Strip ~~~, ~~~~ and ~~~~~ out of signatures @see cleanSigInSig
3802 * 2) Substitute all transclusions
3804 * @param string $text
3805 * @param $parsing Whether we're cleaning (preferences save) or parsing
3806 * @return string Signature text
3808 function cleanSig( $text, $parsing = false ) {
3810 $this->startExternalParse( $wgTitle, new ParserOptions(), $parsing ? OT_WIKI
: OT_MSG
);
3812 $substWord = MagicWord
::get( 'subst' );
3813 $substRegex = '/\{\{(?!(?:' . $substWord->getBaseRegex() . '))/x' . $substWord->getRegexCase();
3814 $substText = '{{' . $substWord->getSynonym( 0 );
3816 $text = preg_replace( $substRegex, $substText, $text );
3817 $text = $this->cleanSigInSig( $text );
3818 $text = $this->replaceVariables( $text );
3820 $this->clearState();
3825 * Strip ~~~, ~~~~ and ~~~~~ out of signatures
3826 * @param string $text
3827 * @return string Signature text with /~{3,5}/ removed
3829 function cleanSigInSig( $text ) {
3830 $text = preg_replace( '/~{3,5}/', '', $text );
3835 * Set up some variables which are usually set up in parse()
3836 * so that an external function can call some class members with confidence
3839 function startExternalParse( &$title, $options, $outputType, $clearState = true ) {
3840 $this->mTitle
=& $title;
3841 $this->mOptions
= $options;
3842 $this->setOutputType( $outputType );
3843 if ( $clearState ) {
3844 $this->clearState();
3849 * Transform a MediaWiki message by replacing magic variables.
3851 * @param string $text the text to transform
3852 * @param ParserOptions $options options
3853 * @return string the text with variables substituted
3856 function transformMsg( $text, $options ) {
3858 static $executing = false;
3860 $fname = "Parser::transformMsg";
3862 # Guard against infinite recursion
3868 wfProfileIn($fname);
3870 if ( $wgTitle && !( $wgTitle instanceof FakeTitle
) ) {
3871 $this->mTitle
= $wgTitle;
3873 $this->mTitle
= Title
::newFromText('msg');
3875 $this->mOptions
= $options;
3876 $this->setOutputType( OT_MSG
);
3877 $this->clearState();
3878 $text = $this->replaceVariables( $text );
3881 wfProfileOut($fname);
3886 * Create an HTML-style tag, e.g. <yourtag>special text</yourtag>
3887 * The callback should have the following form:
3888 * function myParserHook( $text, $params, &$parser ) { ... }
3890 * Transform and return $text. Use $parser for any required context, e.g. use
3891 * $parser->getTitle() and $parser->getOptions() not $wgTitle or $wgOut->mParserOptions
3895 * @param mixed $tag The tag to use, e.g. 'hook' for <hook>
3896 * @param mixed $callback The callback function (and object) to use for the tag
3898 * @return The old value of the mTagHooks array associated with the hook
3900 function setHook( $tag, $callback ) {
3901 $tag = strtolower( $tag );
3902 $oldVal = isset( $this->mTagHooks
[$tag] ) ?
$this->mTagHooks
[$tag] : null;
3903 $this->mTagHooks
[$tag] = $callback;
3909 * Create a function, e.g. {{sum:1|2|3}}
3910 * The callback function should have the form:
3911 * function myParserFunction( &$parser, $arg1, $arg2, $arg3 ) { ... }
3913 * The callback may either return the text result of the function, or an array with the text
3914 * in element 0, and a number of flags in the other elements. The names of the flags are
3915 * specified in the keys. Valid flags are:
3916 * found The text returned is valid, stop processing the template. This
3918 * nowiki Wiki markup in the return value should be escaped
3919 * noparse Unsafe HTML tags should not be stripped, etc.
3920 * noargs Don't replace triple-brace arguments in the return value
3921 * isHTML The returned text is HTML, armour it against wikitext transformation
3925 * @param string $id The magic word ID
3926 * @param mixed $callback The callback function (and object) to use
3927 * @param integer $flags a combination of the following flags:
3928 * SFH_NO_HASH No leading hash, i.e. {{plural:...}} instead of {{#if:...}}
3930 * @return The old callback function for this name, if any
3932 function setFunctionHook( $id, $callback, $flags = 0 ) {
3933 $oldVal = isset( $this->mFunctionHooks
[$id] ) ?
$this->mFunctionHooks
[$id] : null;
3934 $this->mFunctionHooks
[$id] = $callback;
3936 # Add to function cache
3937 $mw = MagicWord
::get( $id );
3939 throw new MWException( 'Parser::setFunctionHook() expecting a magic word identifier.' );
3941 $synonyms = $mw->getSynonyms();
3942 $sensitive = intval( $mw->isCaseSensitive() );
3944 foreach ( $synonyms as $syn ) {
3946 if ( !$sensitive ) {
3947 $syn = strtolower( $syn );
3950 if ( !( $flags & SFH_NO_HASH
) ) {
3953 # Remove trailing colon
3954 if ( substr( $syn, -1, 1 ) == ':' ) {
3955 $syn = substr( $syn, 0, -1 );
3957 $this->mFunctionSynonyms
[$sensitive][$syn] = $id;
3963 * Get all registered function hook identifiers
3967 function getFunctionHooks() {
3968 return array_keys( $this->mFunctionHooks
);
3972 * Replace <!--LINK--> link placeholders with actual links, in the buffer
3973 * Placeholders created in Skin::makeLinkObj()
3974 * Returns an array of links found, indexed by PDBK:
3978 * $options is a bit field, RLH_FOR_UPDATE to select for update
3980 function replaceLinkHolders( &$text, $options = 0 ) {
3984 $fname = 'Parser::replaceLinkHolders';
3985 wfProfileIn( $fname );
3989 $sk = $this->mOptions
->getSkin();
3990 $linkCache =& LinkCache
::singleton();
3992 if ( !empty( $this->mLinkHolders
['namespaces'] ) ) {
3993 wfProfileIn( $fname.'-check' );
3994 $dbr = wfGetDB( DB_SLAVE
);
3995 $page = $dbr->tableName( 'page' );
3996 $threshold = $wgUser->getOption('stubthreshold');
3999 asort( $this->mLinkHolders
['namespaces'] );
4004 foreach ( $this->mLinkHolders
['namespaces'] as $key => $ns ) {
4006 $title = $this->mLinkHolders
['titles'][$key];
4008 # Skip invalid entries.
4009 # Result will be ugly, but prevents crash.
4010 if ( is_null( $title ) ) {
4013 $pdbk = $pdbks[$key] = $title->getPrefixedDBkey();
4015 # Check if it's a static known link, e.g. interwiki
4016 if ( $title->isAlwaysKnown() ) {
4017 $colours[$pdbk] = 1;
4018 } elseif ( ( $id = $linkCache->getGoodLinkID( $pdbk ) ) != 0 ) {
4019 $colours[$pdbk] = 1;
4020 $this->mOutput
->addLink( $title, $id );
4021 } elseif ( $linkCache->isBadLink( $pdbk ) ) {
4022 $colours[$pdbk] = 0;
4024 # Not in the link cache, add it to the query
4025 if ( !isset( $current ) ) {
4027 $query = "SELECT page_id, page_namespace, page_title";
4028 if ( $threshold > 0 ) {
4029 $query .= ', page_len, page_is_redirect';
4031 $query .= " FROM $page WHERE (page_namespace=$ns AND page_title IN(";
4032 } elseif ( $current != $ns ) {
4034 $query .= ")) OR (page_namespace=$ns AND page_title IN(";
4039 $query .= $dbr->addQuotes( $this->mLinkHolders
['dbkeys'][$key] );
4044 if ( $options & RLH_FOR_UPDATE
) {
4045 $query .= ' FOR UPDATE';
4048 $res = $dbr->query( $query, $fname );
4050 # Fetch data and form into an associative array
4051 # non-existent = broken
4054 while ( $s = $dbr->fetchObject($res) ) {
4055 $title = Title
::makeTitle( $s->page_namespace
, $s->page_title
);
4056 $pdbk = $title->getPrefixedDBkey();
4057 $linkCache->addGoodLinkObj( $s->page_id
, $title );
4058 $this->mOutput
->addLink( $title, $s->page_id
);
4060 if ( $threshold > 0 ) {
4061 $size = $s->page_len
;
4062 if ( $s->page_is_redirect ||
$s->page_namespace
!= 0 ||
$size >= $threshold ) {
4063 $colours[$pdbk] = 1;
4065 $colours[$pdbk] = 2;
4068 $colours[$pdbk] = 1;
4072 wfProfileOut( $fname.'-check' );
4074 # Do a second query for different language variants of links and categories
4075 if($wgContLang->hasVariants()){
4076 $linkBatch = new LinkBatch();
4077 $variantMap = array(); // maps $pdbkey_Variant => $keys (of link holders)
4078 $categoryMap = array(); // maps $category_variant => $category (dbkeys)
4079 $varCategories = array(); // category replacements oldDBkey => newDBkey
4081 $categories = $this->mOutput
->getCategoryLinks();
4083 // Add variants of links to link batch
4084 foreach ( $this->mLinkHolders
['namespaces'] as $key => $ns ) {
4085 $title = $this->mLinkHolders
['titles'][$key];
4086 if ( is_null( $title ) )
4089 $pdbk = $title->getPrefixedDBkey();
4090 $titleText = $title->getText();
4092 // generate all variants of the link title text
4093 $allTextVariants = $wgContLang->convertLinkToAllVariants($titleText);
4095 // if link was not found (in first query), add all variants to query
4096 if ( !isset($colours[$pdbk]) ){
4097 foreach($allTextVariants as $textVariant){
4098 if($textVariant != $titleText){
4099 $variantTitle = Title
::makeTitle( $ns, $textVariant );
4100 if(is_null($variantTitle)) continue;
4101 $linkBatch->addObj( $variantTitle );
4102 $variantMap[$variantTitle->getPrefixedDBkey()][] = $key;
4108 // process categories, check if a category exists in some variant
4109 foreach( $categories as $category){
4110 $variants = $wgContLang->convertLinkToAllVariants($category);
4111 foreach($variants as $variant){
4112 if($variant != $category){
4113 $variantTitle = Title
::newFromDBkey( Title
::makeName(NS_CATEGORY
,$variant) );
4114 if(is_null($variantTitle)) continue;
4115 $linkBatch->addObj( $variantTitle );
4116 $categoryMap[$variant] = $category;
4122 if(!$linkBatch->isEmpty()){
4124 $titleClause = $linkBatch->constructSet('page', $dbr);
4126 $variantQuery = "SELECT page_id, page_namespace, page_title";
4127 if ( $threshold > 0 ) {
4128 $variantQuery .= ', page_len, page_is_redirect';
4131 $variantQuery .= " FROM $page WHERE $titleClause";
4132 if ( $options & RLH_FOR_UPDATE
) {
4133 $variantQuery .= ' FOR UPDATE';
4136 $varRes = $dbr->query( $variantQuery, $fname );
4138 // for each found variants, figure out link holders and replace
4139 while ( $s = $dbr->fetchObject($varRes) ) {
4141 $variantTitle = Title
::makeTitle( $s->page_namespace
, $s->page_title
);
4142 $varPdbk = $variantTitle->getPrefixedDBkey();
4143 $vardbk = $variantTitle->getDBkey();
4145 $holderKeys = array();
4146 if(isset($variantMap[$varPdbk])){
4147 $holderKeys = $variantMap[$varPdbk];
4148 $linkCache->addGoodLinkObj( $s->page_id
, $variantTitle );
4149 $this->mOutput
->addLink( $variantTitle, $s->page_id
);
4152 // loop over link holders
4153 foreach($holderKeys as $key){
4154 $title = $this->mLinkHolders
['titles'][$key];
4155 if ( is_null( $title ) ) continue;
4157 $pdbk = $title->getPrefixedDBkey();
4159 if(!isset($colours[$pdbk])){
4160 // found link in some of the variants, replace the link holder data
4161 $this->mLinkHolders
['titles'][$key] = $variantTitle;
4162 $this->mLinkHolders
['dbkeys'][$key] = $variantTitle->getDBkey();
4164 // set pdbk and colour
4165 $pdbks[$key] = $varPdbk;
4166 if ( $threshold > 0 ) {
4167 $size = $s->page_len
;
4168 if ( $s->page_is_redirect ||
$s->page_namespace
!= 0 ||
$size >= $threshold ) {
4169 $colours[$varPdbk] = 1;
4171 $colours[$varPdbk] = 2;
4175 $colours[$varPdbk] = 1;
4180 // check if the object is a variant of a category
4181 if(isset($categoryMap[$vardbk])){
4182 $oldkey = $categoryMap[$vardbk];
4183 if($oldkey != $vardbk)
4184 $varCategories[$oldkey]=$vardbk;
4188 // rebuild the categories in original order (if there are replacements)
4189 if(count($varCategories)>0){
4191 $originalCats = $this->mOutput
->getCategories();
4192 foreach($originalCats as $cat => $sortkey){
4193 // make the replacement
4194 if( array_key_exists($cat,$varCategories) )
4195 $newCats[$varCategories[$cat]] = $sortkey;
4196 else $newCats[$cat] = $sortkey;
4198 $this->mOutput
->setCategoryLinks($newCats);
4203 # Construct search and replace arrays
4204 wfProfileIn( $fname.'-construct' );
4205 $replacePairs = array();
4206 foreach ( $this->mLinkHolders
['namespaces'] as $key => $ns ) {
4207 $pdbk = $pdbks[$key];
4208 $searchkey = "<!--LINK $key-->";
4209 $title = $this->mLinkHolders
['titles'][$key];
4210 if ( empty( $colours[$pdbk] ) ) {
4211 $linkCache->addBadLinkObj( $title );
4212 $colours[$pdbk] = 0;
4213 $this->mOutput
->addLink( $title, 0 );
4214 $replacePairs[$searchkey] = $sk->makeBrokenLinkObj( $title,
4215 $this->mLinkHolders
['texts'][$key],
4216 $this->mLinkHolders
['queries'][$key] );
4217 } elseif ( $colours[$pdbk] == 1 ) {
4218 $replacePairs[$searchkey] = $sk->makeKnownLinkObj( $title,
4219 $this->mLinkHolders
['texts'][$key],
4220 $this->mLinkHolders
['queries'][$key] );
4221 } elseif ( $colours[$pdbk] == 2 ) {
4222 $replacePairs[$searchkey] = $sk->makeStubLinkObj( $title,
4223 $this->mLinkHolders
['texts'][$key],
4224 $this->mLinkHolders
['queries'][$key] );
4227 $replacer = new HashtableReplacer( $replacePairs, 1 );
4228 wfProfileOut( $fname.'-construct' );
4231 wfProfileIn( $fname.'-replace' );
4232 $text = preg_replace_callback(
4233 '/(<!--LINK .*?-->)/',
4237 wfProfileOut( $fname.'-replace' );
4240 # Now process interwiki link holders
4241 # This is quite a bit simpler than internal links
4242 if ( !empty( $this->mInterwikiLinkHolders
['texts'] ) ) {
4243 wfProfileIn( $fname.'-interwiki' );
4244 # Make interwiki link HTML
4245 $replacePairs = array();
4246 foreach( $this->mInterwikiLinkHolders
['texts'] as $key => $link ) {
4247 $title = $this->mInterwikiLinkHolders
['titles'][$key];
4248 $replacePairs[$key] = $sk->makeLinkObj( $title, $link );
4250 $replacer = new HashtableReplacer( $replacePairs, 1 );
4252 $text = preg_replace_callback(
4253 '/<!--IWLINK (.*?)-->/',
4256 wfProfileOut( $fname.'-interwiki' );
4259 wfProfileOut( $fname );
4264 * Replace <!--LINK--> link placeholders with plain text of links
4265 * (not HTML-formatted).
4266 * @param string $text
4269 function replaceLinkHoldersText( $text ) {
4270 $fname = 'Parser::replaceLinkHoldersText';
4271 wfProfileIn( $fname );
4273 $text = preg_replace_callback(
4274 '/<!--(LINK|IWLINK) (.*?)-->/',
4275 array( &$this, 'replaceLinkHoldersTextCallback' ),
4278 wfProfileOut( $fname );
4283 * @param array $matches
4287 function replaceLinkHoldersTextCallback( $matches ) {
4288 $type = $matches[1];
4290 if( $type == 'LINK' ) {
4291 if( isset( $this->mLinkHolders
['texts'][$key] ) ) {
4292 return $this->mLinkHolders
['texts'][$key];
4294 } elseif( $type == 'IWLINK' ) {
4295 if( isset( $this->mInterwikiLinkHolders
['texts'][$key] ) ) {
4296 return $this->mInterwikiLinkHolders
['texts'][$key];
4303 * Tag hook handler for 'pre'.
4305 function renderPreTag( $text, $attribs ) {
4306 // Backwards-compatibility hack
4307 $content = StringUtils
::delimiterReplace( '<nowiki>', '</nowiki>', '$1', $text, 'i' );
4309 $attribs = Sanitizer
::validateTagAttributes( $attribs, 'pre' );
4310 return wfOpenElement( 'pre', $attribs ) .
4311 Xml
::escapeTagsOnly( $content ) .
4316 * Renders an image gallery from a text with one line per image.
4317 * text labels may be given by using |-style alternative text. E.g.
4318 * Image:one.jpg|The number "1"
4319 * Image:tree.jpg|A tree
4320 * given as text will return the HTML of a gallery with two images,
4321 * labeled 'The number "1"' and
4324 function renderImageGallery( $text, $params ) {
4325 $ig = new ImageGallery();
4326 $ig->setContextTitle( $this->mTitle
);
4327 $ig->setShowBytes( false );
4328 $ig->setShowFilename( false );
4330 $ig->useSkin( $this->mOptions
->getSkin() );
4332 if( isset( $params['caption'] ) ) {
4333 $caption = $params['caption'];
4334 $caption = htmlspecialchars( $caption );
4335 $caption = $this->replaceInternalLinks( $caption );
4336 $ig->setCaptionHtml( $caption );
4338 if( isset( $params['perrow'] ) ) {
4339 $ig->setPerRow( $params['perrow'] );
4341 if( isset( $params['widths'] ) ) {
4342 $ig->setWidths( $params['widths'] );
4344 if( isset( $params['heights'] ) ) {
4345 $ig->setHeights( $params['heights'] );
4348 $lines = explode( "\n", $text );
4349 foreach ( $lines as $line ) {
4350 # match lines like these:
4351 # Image:someimage.jpg|This is some image
4353 preg_match( "/^([^|]+)(\\|(.*))?$/", $line, $matches );
4355 if ( count( $matches ) == 0 ) {
4358 $tp = Title
::newFromText( $matches[1] );
4360 if( is_null( $nt ) ) {
4361 # Bogus title. Ignore these so we don't bomb out later.
4364 if ( isset( $matches[3] ) ) {
4365 $label = $matches[3];
4370 $pout = $this->parse( $label,
4373 false, // Strip whitespace...?
4374 false // Don't clear state!
4376 $html = $pout->getText();
4378 $ig->add( new Image( $nt ), $html );
4380 # Only add real images (bug #5586)
4381 if ( $nt->getNamespace() == NS_IMAGE
) {
4382 $this->mOutput
->addImage( $nt->getDBkey() );
4385 return $ig->toHTML();
4389 * Parse image options text and use it to make an image
4391 function makeImage( $nt, $options ) {
4392 # @TODO: let the MediaHandler specify its transform parameters
4394 # Check if the options text is of the form "options|alt text"
4396 # * thumbnail make a thumbnail with enlarge-icon and caption, alignment depends on lang
4397 # * left no resizing, just left align. label is used for alt= only
4398 # * right same, but right aligned
4399 # * none same, but not aligned
4400 # * ___px scale to ___ pixels width, no aligning. e.g. use in taxobox
4401 # * center center the image
4402 # * framed Keep original image size, no magnify-button.
4403 # vertical-align values (no % or length right now):
4414 $part = array_map( 'trim', explode( '|', $options) );
4417 $alignments = array( 'left', 'right', 'center', 'none', 'baseline', 'sub', 'super', 'top', 'text-top', 'middle', 'bottom', 'text-bottom' );
4418 foreach ( $alignments as $alignment ) {
4419 $mwAlign[$alignment] =& MagicWord
::get( 'img_'.$alignment );
4421 $mwThumb =& MagicWord
::get( 'img_thumbnail' );
4422 $mwManualThumb =& MagicWord
::get( 'img_manualthumb' );
4423 $mwWidth =& MagicWord
::get( 'img_width' );
4424 $mwFramed =& MagicWord
::get( 'img_framed' );
4425 $mwPage =& MagicWord
::get( 'img_page' );
4429 $framed = $thumb = false;
4430 $manual_thumb = '' ;
4431 $align = $valign = '';
4432 $sk = $this->mOptions
->getSkin();
4434 foreach( $part as $val ) {
4435 if ( !is_null( $mwThumb->matchVariableStartToEnd($val) ) ) {
4437 } elseif ( ! is_null( $match = $mwManualThumb->matchVariableStartToEnd($val) ) ) {
4438 # use manually specified thumbnail
4440 $manual_thumb = $match;
4442 foreach( $alignments as $alignment ) {
4443 if ( ! is_null( $mwAlign[$alignment]->matchVariableStartToEnd($val) ) ) {
4444 switch ( $alignment ) {
4445 case 'left': case 'right': case 'center': case 'none':
4446 $align = $alignment; break;
4448 $valign = $alignment;
4453 if ( ! is_null( $match = $mwPage->matchVariableStartToEnd($val) ) ) {
4454 # Select a page in a multipage document
4455 $params['page'] = $match;
4456 } elseif ( !isset( $params['width'] ) && ! is_null( $match = $mwWidth->matchVariableStartToEnd($val) ) ) {
4457 wfDebug( "img_width match: $match\n" );
4458 # $match is the image width in pixels
4460 if ( preg_match( '/^([0-9]*)x([0-9]*)$/', $match, $m ) ) {
4461 $params['width'] = intval( $m[1] );
4462 $params['height'] = intval( $m[2] );
4464 $params['width'] = intval($match);
4466 } elseif ( ! is_null( $mwFramed->matchVariableStartToEnd($val) ) ) {
4473 # Strip bad stuff out of the alt text
4474 $alt = $this->replaceLinkHoldersText( $caption );
4476 # make sure there are no placeholders in thumbnail attributes
4477 # that are later expanded to html- so expand them now and
4479 $alt = $this->mStripState
->unstripBoth( $alt );
4480 $alt = Sanitizer
::stripAllTags( $alt );
4482 # Linker does the rest
4483 return $sk->makeImageLinkObj( $nt, $caption, $alt, $align, $params, $framed, $thumb, $manual_thumb, $valign );
4487 * Set a flag in the output object indicating that the content is dynamic and
4488 * shouldn't be cached.
4490 function disableCache() {
4491 wfDebug( "Parser output marked as uncacheable.\n" );
4492 $this->mOutput
->mCacheTime
= -1;
4496 * Callback from the Sanitizer for expanding items found in HTML attribute
4497 * values, so they can be safely tested and escaped.
4498 * @param string $text
4499 * @param array $args
4503 function attributeStripCallback( &$text, $args ) {
4504 $text = $this->replaceVariables( $text, $args );
4505 $text = $this->mStripState
->unstripBoth( $text );
4514 function Title( $x = NULL ) { return wfSetVar( $this->mTitle
, $x ); }
4515 function Options( $x = NULL ) { return wfSetVar( $this->mOptions
, $x ); }
4516 function OutputType( $x = NULL ) { return wfSetVar( $this->mOutputType
, $x ); }
4522 function getTags() { return array_keys( $this->mTagHooks
); }
4527 * Break wikitext input into sections, and either pull or replace
4528 * some particular section's text.
4530 * External callers should use the getSection and replaceSection methods.
4532 * @param $text Page wikitext
4533 * @param $section Numbered section. 0 pulls the text before the first
4534 * heading; other numbers will pull the given section
4535 * along with its lower-level subsections.
4536 * @param $mode One of "get" or "replace"
4537 * @param $newtext Replacement text for section data.
4538 * @return string for "get", the extracted section text.
4539 * for "replace", the whole page with the section replaced.
4541 private function extractSections( $text, $section, $mode, $newtext='' ) {
4542 # strip NOWIKI etc. to avoid confusion (true-parameter causes HTML
4543 # comments to be stripped as well)
4544 $stripState = new StripState
;
4546 $oldOutputType = $this->mOutputType
;
4547 $oldOptions = $this->mOptions
;
4548 $this->mOptions
= new ParserOptions();
4549 $this->setOutputType( OT_WIKI
);
4551 $striptext = $this->strip( $text, $stripState, true );
4553 $this->setOutputType( $oldOutputType );
4554 $this->mOptions
= $oldOptions;
4556 # now that we can be sure that no pseudo-sections are in the source,
4557 # split it up by section
4558 $uniq = preg_quote( $this->uniqPrefix(), '/' );
4559 $comment = "(?:$uniq-!--.*?QINU)";
4564 (?:$comment|<\/?noinclude>)* # Initial comments will be stripped
4565 (=+) # Should this be limited to 6?
4566 .+? # Section title...
4567 \\2 # Ending = count must match start
4568 (?:$comment|<\/?noinclude>|[ \\t]+)* # Trailing whitespace ok
4577 PREG_SPLIT_DELIM_CAPTURE
);
4579 if( $mode == "get" ) {
4580 if( $section == 0 ) {
4581 // "Section 0" returns the content before any other section.
4584 //track missing section, will replace if found.
4587 } elseif( $mode == "replace" ) {
4588 if( $section == 0 ) {
4589 $rv = $newtext . "\n\n";
4598 for( $index = 1; $index < count( $secs ); ) {
4599 $headerLine = $secs[$index++
];
4600 if( $secs[$index] ) {
4602 $headerLevel = strlen( $secs[$index++
] );
4606 $headerLevel = intval( $secs[$index++
] );
4608 $content = $secs[$index++
];
4611 if( $mode == "get" ) {
4612 if( $count == $section ) {
4613 $rv = $headerLine . $content;
4614 $sectionLevel = $headerLevel;
4615 } elseif( $count > $section ) {
4616 if( $sectionLevel && $headerLevel > $sectionLevel ) {
4617 $rv .= $headerLine . $content;
4619 // Broke out to a higher-level section
4623 } elseif( $mode == "replace" ) {
4624 if( $count < $section ) {
4625 $rv .= $headerLine . $content;
4626 } elseif( $count == $section ) {
4627 $rv .= $newtext . "\n\n";
4628 $sectionLevel = $headerLevel;
4629 } elseif( $count > $section ) {
4630 if( $headerLevel <= $sectionLevel ) {
4631 // Passed the section's sub-parts.
4635 $rv .= $headerLine . $content;
4641 # reinsert stripped tags
4642 $rv = trim( $stripState->unstripBoth( $rv ) );
4648 * This function returns the text of a section, specified by a number ($section).
4649 * A section is text under a heading like == Heading == or \<h1\>Heading\</h1\>, or
4650 * the first section before any such heading (section 0).
4652 * If a section contains subsections, these are also returned.
4654 * @param $text String: text to look in
4655 * @param $section Integer: section number
4656 * @param $deftext: default to return if section is not found
4657 * @return string text of the requested section
4659 public function getSection( $text, $section, $deftext='' ) {
4660 return $this->extractSections( $text, $section, "get", $deftext );
4663 public function replaceSection( $oldtext, $section, $text ) {
4664 return $this->extractSections( $oldtext, $section, "replace", $text );
4668 * Get the timestamp associated with the current revision, adjusted for
4669 * the default server-local timestamp
4671 function getRevisionTimestamp() {
4672 if ( is_null( $this->mRevisionTimestamp
) ) {
4673 wfProfileIn( __METHOD__
);
4675 $dbr = wfGetDB( DB_SLAVE
);
4676 $timestamp = $dbr->selectField( 'revision', 'rev_timestamp',
4677 array( 'rev_id' => $this->mRevisionId
), __METHOD__
);
4679 // Normalize timestamp to internal MW format for timezone processing.
4680 // This has the added side-effect of replacing a null value with
4681 // the current time, which gives us more sensible behavior for
4683 $timestamp = wfTimestamp( TS_MW
, $timestamp );
4685 // The cryptic '' timezone parameter tells to use the site-default
4686 // timezone offset instead of the user settings.
4688 // Since this value will be saved into the parser cache, served
4689 // to other users, and potentially even used inside links and such,
4690 // it needs to be consistent for all visitors.
4691 $this->mRevisionTimestamp
= $wgContLang->userAdjust( $timestamp, '' );
4693 wfProfileOut( __METHOD__
);
4695 return $this->mRevisionTimestamp
;
4699 * Mutator for $mDefaultSort
4701 * @param $sort New value
4703 public function setDefaultSort( $sort ) {
4704 $this->mDefaultSort
= $sort;
4708 * Accessor for $mDefaultSort
4709 * Will use the title/prefixed title if none is set
4713 public function getDefaultSort() {
4714 if( $this->mDefaultSort
!== false ) {
4715 return $this->mDefaultSort
;
4717 return $this->mTitle
->getNamespace() == NS_CATEGORY
4718 ?
$this->mTitle
->getText()
4719 : $this->mTitle
->getPrefixedText();
4726 * @todo document, briefly.
4727 * @addtogroup Parser
4729 class OnlyIncludeReplacer
{
4732 function replace( $matches ) {
4733 if ( substr( $matches[1], -1 ) == "\n" ) {
4734 $this->output
.= substr( $matches[1], 0, -1 );
4736 $this->output
.= $matches[1];
4742 * @todo document, briefly.
4743 * @addtogroup Parser
4746 var $general, $nowiki;
4748 function __construct() {
4749 $this->general
= new ReplacementArray
;
4750 $this->nowiki
= new ReplacementArray
;
4753 function unstripGeneral( $text ) {
4754 wfProfileIn( __METHOD__
);
4755 $text = $this->general
->replace( $text );
4756 wfProfileOut( __METHOD__
);
4760 function unstripNoWiki( $text ) {
4761 wfProfileIn( __METHOD__
);
4762 $text = $this->nowiki
->replace( $text );
4763 wfProfileOut( __METHOD__
);
4767 function unstripBoth( $text ) {
4768 wfProfileIn( __METHOD__
);
4769 $text = $this->general
->replace( $text );
4770 $text = $this->nowiki
->replace( $text );
4771 wfProfileOut( __METHOD__
);