3 * @defgroup Parser Parser
7 * File for Parser and related classes
12 * PHP Parser - Processes wiki markup (which uses a more user-friendly
13 * syntax, such as "[[link]]" for making links), and provides a one-way
14 * transformation of that wiki markup it into XHTML output / markup
15 * (which in turn the browser understands, and can display).
18 * There are five main entry points into the Parser class:
20 * produces HTML output
22 * produces altered wiki markup.
24 * removes HTML comments and expands templates
26 * Cleans a signature before saving it to preferences
28 * Extracts sections from an article for section editing
30 * Removes <noinclude> sections, and <includeonly> tags.
33 * objects: $wgLang, $wgContLang
35 * NOT $wgArticle, $wgUser or $wgTitle. Keep them away!
38 * $wgUseTex*, $wgUseDynamicDates*, $wgInterwikiMagic*,
39 * $wgNamespacesWithSubpages, $wgAllowExternalImages*,
40 * $wgLocaltimezone, $wgAllowSpecialInclusion*,
43 * * only within ParserOptions
50 * Update this version number when the ParserOutput format
51 * changes in an incompatible way, so the parser cache
52 * can automatically discard old data.
54 const VERSION
= '1.6.4';
56 # Flags for Parser::setFunctionHook
57 # Also available as global constants from Defines.php
58 const SFH_NO_HASH
= 1;
59 const SFH_OBJECT_ARGS
= 2;
61 # Constants needed for external link processing
62 # Everything except bracket, space, or control characters
63 const EXT_LINK_URL_CLASS
= '[^][<>"\\x00-\\x20\\x7F]';
64 const EXT_IMAGE_REGEX
= '/^(http:\/\/|https:\/\/)([^][<>"\\x00-\\x20\\x7F]+)
65 \\/([A-Za-z0-9_.,~%\\-+&;#*?!=()@\\x80-\\xFF]+)\\.((?i)gif|png|jpg|jpeg)$/Sx';
67 # State constants for the definition list colon extraction
68 const COLON_STATE_TEXT
= 0;
69 const COLON_STATE_TAG
= 1;
70 const COLON_STATE_TAGSTART
= 2;
71 const COLON_STATE_CLOSETAG
= 3;
72 const COLON_STATE_TAGSLASH
= 4;
73 const COLON_STATE_COMMENT
= 5;
74 const COLON_STATE_COMMENTDASH
= 6;
75 const COLON_STATE_COMMENTDASHDASH
= 7;
77 # Flags for preprocessToDom
78 const PTD_FOR_INCLUSION
= 1;
80 # Allowed values for $this->mOutputType
81 # Parameter to startExternalParse().
82 const OT_HTML
= 1; # like parse()
83 const OT_WIKI
= 2; # like preSaveTransform()
84 const OT_PREPROCESS
= 3; # like preprocess()
86 const OT_PLAIN
= 4; # like extractSections() - portions of the original are returned unchanged.
88 # Marker Suffix needs to be accessible staticly.
89 const MARKER_SUFFIX
= "-QINU\x7f";
92 var $mTagHooks, $mTransparentTagHooks, $mFunctionHooks, $mFunctionSynonyms, $mVariables;
93 var $mSubstWords, $mImageParams, $mImageParamsMagicArray, $mStripList, $mMarkerIndex;
94 var $mPreprocessor, $mExtLinkBracketedRegex, $mUrlProtocols, $mDefaultStripList;
95 var $mVarCache, $mConf, $mFunctionTagHooks;
98 # Cleared with clearState():
99 var $mOutput, $mAutonumber, $mDTopen, $mStripState;
100 var $mIncludeCount, $mArgStack, $mLastSection, $mInPre;
101 var $mLinkHolders, $mLinkID;
102 var $mIncludeSizes, $mPPNodeCount, $mDefaultSort;
103 var $mTplExpandCache; # empty-frame expansion cache
104 var $mTplRedirCache, $mTplDomCache, $mHeadings, $mDoubleUnderscores;
105 var $mExpensiveFunctionCount; # number of expensive parser function calls
108 # These are variables reset at least once per parse regardless of $clearState
109 var $mOptions; # ParserOptions object
110 var $mTitle; # Title context, used for self-link rendering and similar things
111 var $mOutputType; # Output type, one of the OT_xxx constants
112 var $ot; # Shortcut alias, see setOutputType()
113 var $mRevisionId; # ID to display in {{REVISIONID}} tags
114 var $mRevisionTimestamp; # The timestamp of the specified revision ID
115 var $mRevIdForTs; # The revision ID which was used to fetch the timestamp
122 function __construct( $conf = array() ) {
123 $this->mConf
= $conf;
124 $this->mTagHooks
= array();
125 $this->mTransparentTagHooks
= array();
126 $this->mFunctionHooks
= array();
127 $this->mFunctionTagHooks
= array();
128 $this->mFunctionSynonyms
= array( 0 => array(), 1 => array() );
129 $this->mDefaultStripList
= $this->mStripList
= array();
130 $this->mUrlProtocols
= wfUrlProtocols();
131 $this->mExtLinkBracketedRegex
= '/\[(\b(' . wfUrlProtocols() . ')'.
132 '[^][<>"\\x00-\\x20\\x7F]+) *([^\]\\x00-\\x08\\x0a-\\x1F]*?)\]/S';
133 $this->mVarCache
= array();
134 if ( isset( $conf['preprocessorClass'] ) ) {
135 $this->mPreprocessorClass
= $conf['preprocessorClass'];
136 } elseif ( extension_loaded( 'domxml' ) ) {
137 # PECL extension that conflicts with the core DOM extension (bug 13770)
138 wfDebug( "Warning: you have the obsolete domxml extension for PHP. Please remove it!\n" );
139 $this->mPreprocessorClass
= 'Preprocessor_Hash';
140 } elseif ( extension_loaded( 'dom' ) ) {
141 $this->mPreprocessorClass
= 'Preprocessor_DOM';
143 $this->mPreprocessorClass
= 'Preprocessor_Hash';
145 $this->mMarkerIndex
= 0;
146 $this->mFirstCall
= true;
150 * Reduce memory usage to reduce the impact of circular references
152 function __destruct() {
153 if ( isset( $this->mLinkHolders
) ) {
154 $this->mLinkHolders
->__destruct();
156 foreach ( $this as $name => $value ) {
157 unset( $this->$name );
162 * Do various kinds of initialisation on the first call of the parser
164 function firstCallInit() {
165 if ( !$this->mFirstCall
) {
168 $this->mFirstCall
= false;
170 wfProfileIn( __METHOD__
);
172 CoreParserFunctions
::register( $this );
173 CoreTagHooks
::register( $this );
174 $this->initialiseVariables();
176 wfRunHooks( 'ParserFirstCallInit', array( &$this ) );
177 wfProfileOut( __METHOD__
);
185 function clearState() {
186 wfProfileIn( __METHOD__
);
187 if ( $this->mFirstCall
) {
188 $this->firstCallInit();
190 $this->mOutput
= new ParserOutput
;
191 $this->mAutonumber
= 0;
192 $this->mLastSection
= '';
193 $this->mDTopen
= false;
194 $this->mIncludeCount
= array();
195 $this->mStripState
= new StripState
;
196 $this->mArgStack
= false;
197 $this->mInPre
= false;
198 $this->mLinkHolders
= new LinkHolderArray( $this );
200 $this->mRevisionTimestamp
= $this->mRevisionId
= null;
201 $this->mVarCache
= array();
204 * Prefix for temporary replacement strings for the multipass parser.
205 * \x07 should never appear in input as it's disallowed in XML.
206 * Using it at the front also gives us a little extra robustness
207 * since it shouldn't match when butted up against identifier-like
210 * Must not consist of all title characters, or else it will change
211 * the behaviour of <nowiki> in a link.
213 # $this->mUniqPrefix = "\x07UNIQ" . Parser::getRandomString();
214 # Changed to \x7f to allow XML double-parsing -- TS
215 $this->mUniqPrefix
= "\x7fUNIQ" . self
::getRandomString();
218 # Clear these on every parse, bug 4549
219 $this->mTplExpandCache
= $this->mTplRedirCache
= $this->mTplDomCache
= array();
221 $this->mShowToc
= true;
222 $this->mForceTocPosition
= false;
223 $this->mIncludeSizes
= array(
227 $this->mPPNodeCount
= 0;
228 $this->mDefaultSort
= false;
229 $this->mHeadings
= array();
230 $this->mDoubleUnderscores
= array();
231 $this->mExpensiveFunctionCount
= 0;
234 if ( isset( $this->mPreprocessor
) && $this->mPreprocessor
->parser
!== $this ) {
235 $this->mPreprocessor
= null;
238 wfRunHooks( 'ParserClearState', array( &$this ) );
239 wfProfileOut( __METHOD__
);
243 * Convert wikitext to HTML
244 * Do not call this function recursively.
246 * @param $text String: text we want to parse
247 * @param $title A title object
248 * @param $options ParserOptions
249 * @param $linestart boolean
250 * @param $clearState boolean
251 * @param $revid Int: number to pass in {{REVISIONID}}
252 * @return ParserOutput a ParserOutput
254 public function parse( $text, Title
$title, ParserOptions
$options, $linestart = true, $clearState = true, $revid = null ) {
256 * First pass--just handle <nowiki> sections, pass the rest off
257 * to internalParse() which does all the real work.
260 global $wgUseTidy, $wgAlwaysUseTidy, $wgContLang, $wgDisableLangConversion, $wgDisableTitleConversion;
261 $fname = __METHOD__
.'-' . wfGetCaller();
262 wfProfileIn( __METHOD__
);
263 wfProfileIn( $fname );
269 $options->resetUsage();
270 $this->mOptions
= $options;
271 $this->setTitle( $title ); # Page title has to be set for the pre-processor
273 $oldRevisionId = $this->mRevisionId
;
274 $oldRevisionTimestamp = $this->mRevisionTimestamp
;
275 if ( $revid !== null ) {
276 $this->mRevisionId
= $revid;
277 $this->mRevisionTimestamp
= null;
279 $this->setOutputType( self
::OT_HTML
);
280 wfRunHooks( 'ParserBeforeStrip', array( &$this, &$text, &$this->mStripState
) );
282 wfRunHooks( 'ParserAfterStrip', array( &$this, &$text, &$this->mStripState
) );
283 $text = $this->internalParse( $text );
285 $text = $this->mStripState
->unstripGeneral( $text );
287 # Clean up special characters, only run once, next-to-last before doBlockLevels
289 # french spaces, last one Guillemet-left
290 # only if there is something before the space
291 '/(.) (?=\\?|:|;|!|%|\\302\\273)/' => '\\1 \\2',
292 # french spaces, Guillemet-right
293 '/(\\302\\253) /' => '\\1 ',
294 '/ (!\s*important)/' => ' \\1', # Beware of CSS magic word !important, bug #11874.
296 $text = preg_replace( array_keys( $fixtags ), array_values( $fixtags ), $text );
298 $text = $this->doBlockLevels( $text, $linestart );
300 $this->replaceLinkHolders( $text );
303 * The page doesn't get language converted if
305 * b) Content isn't converted
306 * c) It's a conversion table
308 if ( !( $wgDisableLangConversion
309 ||
isset( $this->mDoubleUnderscores
['nocontentconvert'] )
310 ||
$this->mTitle
->isConversionTable() ) ) {
312 # The position of the convert() call should not be changed. it
313 # assumes that the links are all replaced and the only thing left
314 # is the <nowiki> mark.
316 $text = $wgContLang->convert( $text );
320 * A page get its title converted except:
321 * a) Language conversion is globally disabled
322 * b) Title convert is globally disabled
323 * c) The page is a redirect page
324 * d) User request with a "linkconvert" set to "no"
325 * e) A "nocontentconvert" magic word has been set
326 * f) A "notitleconvert" magic word has been set
327 * g) User sets "noconvertlink" in his/her preference
329 * Note that if a user tries to set a title in a conversion
330 * rule but content conversion was not done, then the parser
331 * won't pick it up. This is probably expected behavior.
333 if ( !( $wgDisableLangConversion
334 ||
$wgDisableTitleConversion
335 ||
isset( $this->mDoubleUnderscores
['nocontentconvert'] )
336 ||
isset( $this->mDoubleUnderscores
['notitleconvert'] )
337 ||
$this->mOutput
->getDisplayTitle() !== false ) )
339 $convruletitle = $wgContLang->getConvRuleTitle();
340 if ( $convruletitle ) {
341 $this->mOutput
->setTitleText( $convruletitle );
343 $titleText = $wgContLang->convertTitle( $title );
344 $this->mOutput
->setTitleText( $titleText );
348 $text = $this->mStripState
->unstripNoWiki( $text );
350 wfRunHooks( 'ParserBeforeTidy', array( &$this, &$text ) );
352 //!JF Move to its own function
354 $uniq_prefix = $this->mUniqPrefix
;
356 $elements = array_keys( $this->mTransparentTagHooks
);
357 $text = self
::extractTagsAndParams( $elements, $text, $matches, $uniq_prefix );
359 foreach ( $matches as $marker => $data ) {
360 list( $element, $content, $params, $tag ) = $data;
361 $tagName = strtolower( $element );
362 if ( isset( $this->mTransparentTagHooks
[$tagName] ) ) {
363 $output = call_user_func_array( $this->mTransparentTagHooks
[$tagName], array( $content, $params, $this ) );
367 $this->mStripState
->general
->setPair( $marker, $output );
369 $text = $this->mStripState
->unstripGeneral( $text );
371 $text = Sanitizer
::normalizeCharReferences( $text );
373 if ( ( $wgUseTidy && $this->mOptions
->getTidy() ) ||
$wgAlwaysUseTidy ) {
374 $text = MWTidy
::tidy( $text );
376 # attempt to sanitize at least some nesting problems
377 # (bug #2702 and quite a few others)
379 # ''Something [http://www.cool.com cool''] -->
380 # <i>Something</i><a href="http://www.cool.com"..><i>cool></i></a>
381 '/(<([bi])>)(<([bi])>)?([^<]*)(<\/?a[^<]*>)([^<]*)(<\/\\4>)?(<\/\\2>)/' =>
382 '\\1\\3\\5\\8\\9\\6\\1\\3\\7\\8\\9',
383 # fix up an anchor inside another anchor, only
384 # at least for a single single nested link (bug 3695)
385 '/(<a[^>]+>)([^<]*)(<a[^>]+>[^<]*)<\/a>(.*)<\/a>/' =>
386 '\\1\\2</a>\\3</a>\\1\\4</a>',
387 # fix div inside inline elements- doBlockLevels won't wrap a line which
388 # contains a div, so fix it up here; replace
389 # div with escaped text
390 '/(<([aib]) [^>]+>)([^<]*)(<div([^>]*)>)(.*)(<\/div>)([^<]*)(<\/\\2>)/' =>
391 '\\1\\3<div\\5>\\6</div>\\8\\9',
392 # remove empty italic or bold tag pairs, some
393 # introduced by rules above
394 '/<([bi])><\/\\1>/' => '',
397 $text = preg_replace(
398 array_keys( $tidyregs ),
399 array_values( $tidyregs ),
402 global $wgExpensiveParserFunctionLimit;
403 if ( $this->mExpensiveFunctionCount
> $wgExpensiveParserFunctionLimit ) {
404 $this->limitationWarn( 'expensive-parserfunction', $this->mExpensiveFunctionCount
, $wgExpensiveParserFunctionLimit );
407 wfRunHooks( 'ParserAfterTidy', array( &$this, &$text ) );
409 # Information on include size limits, for the benefit of users who try to skirt them
410 if ( $this->mOptions
->getEnableLimitReport() ) {
411 $max = $this->mOptions
->getMaxIncludeSize();
412 $PFreport = "Expensive parser function count: {$this->mExpensiveFunctionCount}/$wgExpensiveParserFunctionLimit\n";
414 "NewPP limit report\n" .
415 "Preprocessor node count: {$this->mPPNodeCount}/{$this->mOptions->getMaxPPNodeCount()}\n" .
416 "Post-expand include size: {$this->mIncludeSizes['post-expand']}/$max bytes\n" .
417 "Template argument size: {$this->mIncludeSizes['arg']}/$max bytes\n".
419 wfRunHooks( 'ParserLimitReport', array( $this, &$limitReport ) );
420 $text .= "\n<!-- \n$limitReport-->\n";
422 $this->mOutput
->setText( $text );
424 $this->mRevisionId
= $oldRevisionId;
425 $this->mRevisionTimestamp
= $oldRevisionTimestamp;
426 wfProfileOut( $fname );
427 wfProfileOut( __METHOD__
);
429 return $this->mOutput
;
433 * Recursive parser entry point that can be called from an extension tag
436 * If $frame is not provided, then template variables (e.g., {{{1}}}) within $text are not expanded
438 * @param $text String: text extension wants to have parsed
439 * @param $frame PPFrame: The frame to use for expanding any template variables
441 function recursiveTagParse( $text, $frame=false ) {
442 wfProfileIn( __METHOD__
);
443 wfRunHooks( 'ParserBeforeStrip', array( &$this, &$text, &$this->mStripState
) );
444 wfRunHooks( 'ParserAfterStrip', array( &$this, &$text, &$this->mStripState
) );
445 $text = $this->internalParse( $text, false, $frame );
446 wfProfileOut( __METHOD__
);
451 * Expand templates and variables in the text, producing valid, static wikitext.
452 * Also removes comments.
454 function preprocess( $text, $title, $options, $revid = null ) {
455 wfProfileIn( __METHOD__
);
457 $this->setOutputType( self
::OT_PREPROCESS
);
458 $options->resetUsage();
459 $this->mOptions
= $options;
460 $this->setTitle( $title );
461 if ( $revid !== null ) {
462 $this->mRevisionId
= $revid;
464 wfRunHooks( 'ParserBeforeStrip', array( &$this, &$text, &$this->mStripState
) );
465 wfRunHooks( 'ParserAfterStrip', array( &$this, &$text, &$this->mStripState
) );
466 $text = $this->replaceVariables( $text );
467 $text = $this->mStripState
->unstripBoth( $text );
468 wfProfileOut( __METHOD__
);
473 * Process the wikitext for the ?preload= feature. (bug 5210)
475 * <noinclude>, <includeonly> etc. are parsed as for template transclusion,
476 * comments, templates, arguments, tags hooks and parser functions are untouched.
478 public function getPreloadText( $text, $title, $options ) {
479 # Parser (re)initialisation
481 $this->setOutputType( self
::OT_PLAIN
);
482 $options->resetUsage();
483 $this->mOptions
= $options;
484 $this->setTitle( $title );
486 $flags = PPFrame
::NO_ARGS | PPFrame
::NO_TEMPLATES
;
487 $dom = $this->preprocessToDom( $text, self
::PTD_FOR_INCLUSION
);
488 return $this->getPreprocessor()->newFrame()->expand( $dom, $flags );
492 * Get a random string
497 static function getRandomString() {
498 return dechex( mt_rand( 0, 0x7fffffff ) ) . dechex( mt_rand( 0, 0x7fffffff ) );
502 * Accessor for mUniqPrefix.
506 public function uniqPrefix() {
507 if ( !isset( $this->mUniqPrefix
) ) {
508 # @todo Fixme: this is probably *horribly wrong*
509 # LanguageConverter seems to want $wgParser's uniqPrefix, however
510 # if this is called for a parser cache hit, the parser may not
511 # have ever been initialized in the first place.
512 # Not really sure what the heck is supposed to be going on here.
514 # throw new MWException( "Accessing uninitialized mUniqPrefix" );
516 return $this->mUniqPrefix
;
520 * Set the context title
522 function setTitle( $t ) {
523 if ( !$t ||
$t instanceof FakeTitle
) {
524 $t = Title
::newFromText( 'NO TITLE' );
527 if ( strval( $t->getFragment() ) !== '' ) {
528 # Strip the fragment to avoid various odd effects
529 $this->mTitle
= clone $t;
530 $this->mTitle
->setFragment( '' );
537 * Accessor for the Title object
539 * @return Title object
541 function &getTitle() {
542 return $this->mTitle
;
546 * Accessor/mutator for the Title object
548 * @param $x New Title object or null to just get the current one
549 * @return Title object
551 function Title( $x = null ) {
552 return wfSetVar( $this->mTitle
, $x );
556 * Set the output type
558 * @param $ot Integer: new value
560 function setOutputType( $ot ) {
561 $this->mOutputType
= $ot;
564 'html' => $ot == self
::OT_HTML
,
565 'wiki' => $ot == self
::OT_WIKI
,
566 'pre' => $ot == self
::OT_PREPROCESS
,
567 'plain' => $ot == self
::OT_PLAIN
,
572 * Accessor/mutator for the output type
574 * @param $x New value or null to just get the current one
577 function OutputType( $x = null ) {
578 return wfSetVar( $this->mOutputType
, $x );
582 * Get the ParserOutput object
584 * @return ParserOutput object
586 function getOutput() {
587 return $this->mOutput
;
591 * Get the ParserOptions object
593 * @return ParserOptions object
595 function getOptions() {
596 return $this->mOptions
;
600 * Accessor/mutator for the ParserOptions object
602 * @param $x New value or null to just get the current one
603 * @return Current ParserOptions object
605 function Options( $x = null ) {
606 return wfSetVar( $this->mOptions
, $x );
609 function nextLinkID() {
610 return $this->mLinkID++
;
613 function getFunctionLang() {
614 global $wgLang, $wgContLang;
616 $target = $this->mOptions
->getTargetLanguage();
617 if ( $target !== null ) {
620 return $this->mOptions
->getInterfaceMessage() ?
$wgLang : $wgContLang;
625 * Get a preprocessor object
627 * @return Preprocessor instance
629 function getPreprocessor() {
630 if ( !isset( $this->mPreprocessor
) ) {
631 $class = $this->mPreprocessorClass
;
632 $this->mPreprocessor
= new $class( $this );
634 return $this->mPreprocessor
;
638 * Replaces all occurrences of HTML-style comments and the given tags
639 * in the text with a random marker and returns the next text. The output
640 * parameter $matches will be an associative array filled with data in
642 * 'UNIQ-xxxxx' => array(
645 * array( 'param' => 'x' ),
646 * '<element param="x">tag content</element>' ) )
648 * @param $elements list of element names. Comments are always extracted.
649 * @param $text Source text string.
650 * @param $matches Out parameter, Array: extarcted tags
651 * @param $uniq_prefix
652 * @return String: stripped text
656 public function extractTagsAndParams( $elements, $text, &$matches, $uniq_prefix = '' ) {
661 $taglist = implode( '|', $elements );
662 $start = "/<($taglist)(\\s+[^>]*?|\\s*?)(\/?" . ">)|<(!--)/i";
664 while ( $text != '' ) {
665 $p = preg_split( $start, $text, 2, PREG_SPLIT_DELIM_CAPTURE
);
667 if ( count( $p ) < 5 ) {
670 if ( count( $p ) > 5 ) {
684 $marker = "$uniq_prefix-$element-" . sprintf( '%08X', $n++
) . self
::MARKER_SUFFIX
;
685 $stripped .= $marker;
687 if ( $close === '/>' ) {
688 # Empty element tag, <tag />
693 if ( $element === '!--' ) {
696 $end = "/(<\\/$element\\s*>)/i";
698 $q = preg_split( $end, $inside, 2, PREG_SPLIT_DELIM_CAPTURE
);
700 if ( count( $q ) < 3 ) {
701 # No end tag -- let it run out to the end of the text.
710 $matches[$marker] = array( $element,
712 Sanitizer
::decodeTagAttributes( $attributes ),
713 "<$element$attributes$close$content$tail" );
719 * Get a list of strippable XML-like elements
721 function getStripList() {
722 return $this->mStripList
;
726 * @deprecated use replaceVariables
728 function strip( $text, $state, $stripcomments = false , $dontstrip = array() ) {
733 * Restores pre, math, and other extensions removed by strip()
735 * always call unstripNoWiki() after this one
737 * @deprecated use $this->mStripState->unstrip()
739 function unstrip( $text, $state ) {
740 return $state->unstripGeneral( $text );
744 * Always call this after unstrip() to preserve the order
747 * @deprecated use $this->mStripState->unstrip()
749 function unstripNoWiki( $text, $state ) {
750 return $state->unstripNoWiki( $text );
754 * @deprecated use $this->mStripState->unstripBoth()
756 function unstripForHTML( $text ) {
757 return $this->mStripState
->unstripBoth( $text );
761 * Add an item to the strip state
762 * Returns the unique tag which must be inserted into the stripped text
763 * The tag will be replaced with the original text in unstrip()
767 function insertStripItem( $text ) {
768 $rnd = "{$this->mUniqPrefix}-item-{$this->mMarkerIndex}-" . self
::MARKER_SUFFIX
;
769 $this->mMarkerIndex++
;
770 $this->mStripState
->general
->setPair( $rnd, $text );
775 * Interface with html tidy
776 * @deprecated Use MWTidy::tidy()
778 public static function tidy( $text ) {
779 wfDeprecated( __METHOD__
);
780 return MWTidy
::tidy( $text );
784 * parse the wiki syntax used to render tables
788 function doTableStuff( $text ) {
789 wfProfileIn( __METHOD__
);
791 $lines = StringUtils
::explode( "\n", $text );
794 $td_history = array(); # Is currently a td tag open?
795 $last_tag_history = array(); # Save history of last lag activated (td, th or caption)
796 $tr_history = array(); # Is currently a tr tag open?
797 $tr_attributes = array(); # history of tr attributes
798 $has_opened_tr = array(); # Did this table open a <tr> element?
799 $indent_level = 0; # indent level of the table
801 $table_tag = 'table';
805 $caption_tag = 'caption';
807 $extra_table_attribs = null;
808 $extra_tr_attribs = null;
809 $extra_td_attribs = null;
811 $convert_style = false;
813 foreach ( $lines as $outLine ) {
814 $line = trim( $outLine );
816 if ( $line === '' ) { # empty line, go to next line
817 $out .= $outLine."\n";
821 $first_character = $line[0];
824 if ( preg_match( '/^(:*)\{\|(.*)$/', $line , $matches ) ) {
825 # First check if we are starting a new table
826 $indent_level = strlen( $matches[1] );
828 $attributes = $this->mStripState
->unstripBoth( $matches[2] );
830 $attr = Sanitizer
::decodeTagAttributes( $attributes );
832 $mode = @$attr['mode'];
833 if ( !$mode ) $mode = 'data';
835 if ( $mode == 'grid' ||
$mode == 'layout' ) {
840 $caption_tag = 'div';
842 $extra_table_attribs = array( 'class' => 'grid-table' );
843 $extra_tr_attribs = array( 'class' => 'grid-row' );
844 $extra_td_attribs = array( 'class' => 'grid-cell' );
846 $convert_style = true;
849 if ($convert_style) $attr['style'] = Sanitizer
::styleFromAttributes( $attr );
850 $attr = Sanitizer
::validateTagAttributes( $attr, $table_tag );
851 $attributes = Sanitizer
::collapseTagAttributes( $attr, $extra_table_attribs );
853 $outLine = str_repeat( '<dl><dd>' , $indent_level ) . "<$table_tag{$attributes}>";
854 array_push( $td_history , false );
855 array_push( $last_tag_history , '' );
856 array_push( $tr_history , false );
857 array_push( $tr_attributes , '' );
858 array_push( $has_opened_tr , false );
859 } elseif ( count( $td_history ) == 0 ) {
860 # Don't do any of the following
861 $out .= $outLine."\n";
863 } elseif ( substr( $line , 0 , 2 ) === '|}' ) {
864 # We are ending a table
865 $line = "</$table_tag>" . substr( $line , 2 );
866 $last_tag = array_pop( $last_tag_history );
868 if ( !array_pop( $has_opened_tr ) ) {
869 $line = "<$tr_tag><$td_tag></$td_tag></$tr_tag>{$line}";
872 if ( array_pop( $tr_history ) ) {
873 $line = "</$tr_tag>{$line}";
876 if ( array_pop( $td_history ) ) {
877 $line = "</{$last_tag}>{$line}";
879 array_pop( $tr_attributes );
880 $outLine = $line . str_repeat( '</dd></dl>' , $indent_level );
881 } elseif ( substr( $line , 0 , 2 ) === '|-' ) {
882 # Now we have a table row
883 $line = preg_replace( '#^\|-+#', '', $line );
885 # Whats after the tag is now only attributes
886 $attributes = $this->mStripState
->unstripBoth( $line );
888 $attr = Sanitizer
::decodeTagAttributes( $attributes );
889 if ($convert_style) $attr['style'] = Sanitizer
::styleFromAttributes( $attr );
890 $attr = Sanitizer
::validateTagAttributes( $attr, $tr_tag );
891 $attributes = Sanitizer
::collapseTagAttributes( $attr, $extra_tr_attribs );
893 array_pop( $tr_attributes );
894 array_push( $tr_attributes, $attributes );
897 $last_tag = array_pop( $last_tag_history );
898 array_pop( $has_opened_tr );
899 array_push( $has_opened_tr , true );
901 if ( array_pop( $tr_history ) ) {
902 $line = "</$tr_tag>";
905 if ( array_pop( $td_history ) ) {
906 $line = "</{$last_tag}>{$line}";
910 array_push( $tr_history , false );
911 array_push( $td_history , false );
912 array_push( $last_tag_history , '' );
913 } elseif ( $first_character === '|' ||
$first_character === '!' ||
substr( $line , 0 , 2 ) === '|+' ) {
914 # This might be cell elements, td, th or captions
915 if ( substr( $line , 0 , 2 ) === '|+' ) {
916 $first_character = '+';
917 $line = substr( $line , 1 );
920 $line = substr( $line , 1 );
922 if ( $first_character === '!' ) {
923 $line = str_replace( '!!' , '||' , $line );
926 # Split up multiple cells on the same line.
927 # FIXME : This can result in improper nesting of tags processed
928 # by earlier parser steps, but should avoid splitting up eg
929 # attribute values containing literal "||".
930 $cells = StringUtils
::explodeMarkup( '||' , $line );
934 # Loop through each table cell
935 foreach ( $cells as $cell ) {
937 if ( $first_character !== '+' ) {
938 $tr_after = array_pop( $tr_attributes );
939 if ( !array_pop( $tr_history ) ) {
940 $previous = "<$tr_tag{$tr_after}>\n";
942 array_push( $tr_history , true );
943 array_push( $tr_attributes , '' );
944 array_pop( $has_opened_tr );
945 array_push( $has_opened_tr , true );
948 $last_tag = array_pop( $last_tag_history );
950 if ( array_pop( $td_history ) ) {
951 $previous = "</{$last_tag}>\n{$previous}";
954 if ( $first_character === '|' ) {
956 } elseif ( $first_character === '!' ) {
958 } elseif ( $first_character === '+' ) {
959 $last_tag = $caption_tag;
964 array_push( $last_tag_history , $last_tag );
966 # A cell could contain both parameters and data
967 $cell_data = explode( '|' , $cell , 2 );
971 # Bug 553: Note that a '|' inside an invalid link should not
972 # be mistaken as delimiting cell parameters
973 if ( strpos( $cell_data[0], '[[' ) !== false ) {
974 if ($extra_td_attribs) $attributes = Sanitizer
::collapseTagAttributes( $extra_td_attribs );
975 $cell = "{$previous}<{$last_tag}{$attributes}>{$cell}";
976 } elseif ( count( $cell_data ) == 1 ) {
977 if ($extra_td_attribs) $attributes = Sanitizer
::collapseTagAttributes( $extra_td_attribs );
978 $cell = "{$previous}<{$last_tag}{$attributes}>{$cell_data[0]}";
980 $attributes = $this->mStripState
->unstripBoth( $cell_data[0] );
982 $attr = Sanitizer
::decodeTagAttributes( $attributes );
983 if ($convert_style) $attr['style'] = Sanitizer
::styleFromAttributes( $attr );
984 $attr = Sanitizer
::validateTagAttributes( $attr, $last_tag );
985 $attributes = Sanitizer
::collapseTagAttributes( $attr, $extra_td_attribs );
987 $cell = "{$previous}<{$last_tag}{$attributes}>{$cell_data[1]}";
991 array_push( $td_history , true );
994 $out .= $outLine . "\n";
997 # Closing open td, tr && table
998 while ( count( $td_history ) > 0 ) {
999 if ( array_pop( $td_history ) ) {
1000 $out .= "</$td_tag>\n";
1002 if ( array_pop( $tr_history ) ) {
1003 $out .= "</$tr_tag>\n";
1005 if ( !array_pop( $has_opened_tr ) ) {
1006 $out .= "<$tr_tag><$td_tag></$td_tag></$tr_tag>\n" ;
1009 $out .= "</$table_tag>\n";
1012 # Remove trailing line-ending (b/c)
1013 if ( substr( $out, -1 ) === "\n" ) {
1014 $out = substr( $out, 0, -1 );
1017 # special case: don't return empty table
1018 if ( $out === "<$table_tag>\n<$tr_tag><$td_tag></$td_tag></$tr_tag>\n</$table_tag>" ) {
1022 wfProfileOut( __METHOD__
);
1028 * Helper function for parse() that transforms wiki markup into
1029 * HTML. Only called for $mOutputType == self::OT_HTML.
1033 function internalParse( $text, $isMain = true, $frame=false ) {
1034 wfProfileIn( __METHOD__
);
1038 # Hook to suspend the parser in this state
1039 if ( !wfRunHooks( 'ParserBeforeInternalParse', array( &$this, &$text, &$this->mStripState
) ) ) {
1040 wfProfileOut( __METHOD__
);
1044 # if $frame is provided, then use $frame for replacing any variables
1046 # use frame depth to infer how include/noinclude tags should be handled
1047 # depth=0 means this is the top-level document; otherwise it's an included document
1048 if ( !$frame->depth
) {
1051 $flag = Parser
::PTD_FOR_INCLUSION
;
1053 $dom = $this->preprocessToDom( $text, $flag );
1054 $text = $frame->expand( $dom );
1056 # if $frame is not provided, then use old-style replaceVariables
1057 $text = $this->replaceVariables( $text );
1060 $text = Sanitizer
::removeHTMLtags( $text, array( &$this, 'attributeStripCallback' ), false, array_keys( $this->mTransparentTagHooks
) );
1061 wfRunHooks( 'InternalParseBeforeLinks', array( &$this, &$text, &$this->mStripState
) );
1063 # Tables need to come after variable replacement for things to work
1064 # properly; putting them before other transformations should keep
1065 # exciting things like link expansions from showing up in surprising
1067 $text = $this->doTableStuff( $text );
1069 $text = preg_replace( '/(^|\n)-----*/', '\\1<hr />', $text );
1071 $text = $this->doDoubleUnderscore( $text );
1073 $text = $this->doHeadings( $text );
1074 if ( $this->mOptions
->getUseDynamicDates() ) {
1075 $df = DateFormatter
::getInstance();
1076 $text = $df->reformat( $this->mOptions
->getDateFormat(), $text );
1078 $text = $this->replaceInternalLinks( $text );
1079 $text = $this->doAllQuotes( $text );
1080 $text = $this->replaceExternalLinks( $text );
1082 # replaceInternalLinks may sometimes leave behind
1083 # absolute URLs, which have to be masked to hide them from replaceExternalLinks
1084 $text = str_replace( $this->mUniqPrefix
.'NOPARSE', '', $text );
1086 $text = $this->doMagicLinks( $text );
1087 $text = $this->formatHeadings( $text, $origText, $isMain );
1089 wfProfileOut( __METHOD__
);
1094 * Replace special strings like "ISBN xxx" and "RFC xxx" with
1095 * magic external links.
1100 function doMagicLinks( $text ) {
1101 wfProfileIn( __METHOD__
);
1102 $prots = $this->mUrlProtocols
;
1103 $urlChar = self
::EXT_LINK_URL_CLASS
;
1104 $text = preg_replace_callback(
1106 (<a[ \t\r\n>].*?</a>) | # m[1]: Skip link text
1107 (<.*?>) | # m[2]: Skip stuff inside HTML elements' . "
1108 (\\b(?:$prots)$urlChar+) | # m[3]: Free external links" . '
1109 (?:RFC|PMID)\s+([0-9]+) | # m[4]: RFC or PMID, capture number
1110 ISBN\s+(\b # m[5]: ISBN, capture number
1111 (?: 97[89] [\ \-]? )? # optional 13-digit ISBN prefix
1112 (?: [0-9] [\ \-]? ){9} # 9 digits with opt. delimiters
1113 [0-9Xx] # check digit
1115 )!x', array( &$this, 'magicLinkCallback' ), $text );
1116 wfProfileOut( __METHOD__
);
1120 function magicLinkCallback( $m ) {
1121 if ( isset( $m[1] ) && $m[1] !== '' ) {
1124 } elseif ( isset( $m[2] ) && $m[2] !== '' ) {
1127 } elseif ( isset( $m[3] ) && $m[3] !== '' ) {
1128 # Free external link
1129 return $this->makeFreeExternalLink( $m[0] );
1130 } elseif ( isset( $m[4] ) && $m[4] !== '' ) {
1133 if ( substr( $m[0], 0, 3 ) === 'RFC' ) {
1136 $CssClass = 'mw-magiclink-rfc';
1138 } elseif ( substr( $m[0], 0, 4 ) === 'PMID' ) {
1140 $urlmsg = 'pubmedurl';
1141 $CssClass = 'mw-magiclink-pmid';
1144 throw new MWException( __METHOD__
.': unrecognised match type "' .
1145 substr( $m[0], 0, 20 ) . '"' );
1147 $url = wfMsgForContent( $urlmsg, $id);
1148 $sk = $this->mOptions
->getSkin( $this->mTitle
);
1149 $la = $sk->getExternalLinkAttributes( "external $CssClass" );
1150 return "<a href=\"{$url}\"{$la}>{$keyword} {$id}</a>";
1151 } elseif ( isset( $m[5] ) && $m[5] !== '' ) {
1154 $num = strtr( $isbn, array(
1159 $titleObj = SpecialPage
::getTitleFor( 'Booksources', $num );
1161 $titleObj->escapeLocalUrl() .
1162 "\" class=\"internal mw-magiclink-isbn\">ISBN $isbn</a>";
1169 * Make a free external link, given a user-supplied URL
1173 function makeFreeExternalLink( $url ) {
1175 wfProfileIn( __METHOD__
);
1177 $sk = $this->mOptions
->getSkin( $this->mTitle
);
1180 # The characters '<' and '>' (which were escaped by
1181 # removeHTMLtags()) should not be included in
1182 # URLs, per RFC 2396.
1184 if ( preg_match( '/&(lt|gt);/', $url, $m2, PREG_OFFSET_CAPTURE
) ) {
1185 $trail = substr( $url, $m2[0][1] ) . $trail;
1186 $url = substr( $url, 0, $m2[0][1] );
1189 # Move trailing punctuation to $trail
1191 # If there is no left bracket, then consider right brackets fair game too
1192 if ( strpos( $url, '(' ) === false ) {
1196 $numSepChars = strspn( strrev( $url ), $sep );
1197 if ( $numSepChars ) {
1198 $trail = substr( $url, -$numSepChars ) . $trail;
1199 $url = substr( $url, 0, -$numSepChars );
1202 $url = Sanitizer
::cleanUrl( $url );
1204 # Is this an external image?
1205 $text = $this->maybeMakeExternalImage( $url );
1206 if ( $text === false ) {
1207 # Not an image, make a link
1208 $text = $sk->makeExternalLink( $url, $wgContLang->markNoConversion($url), true, 'free',
1209 $this->getExternalLinkAttribs( $url ) );
1210 # Register it in the output object...
1211 # Replace unnecessary URL escape codes with their equivalent characters
1212 $pasteurized = self
::replaceUnusualEscapes( $url );
1213 $this->mOutput
->addExternalLink( $pasteurized );
1215 wfProfileOut( __METHOD__
);
1216 return $text . $trail;
1221 * Parse headers and return html
1225 function doHeadings( $text ) {
1226 wfProfileIn( __METHOD__
);
1227 for ( $i = 6; $i >= 1; --$i ) {
1228 $h = str_repeat( '=', $i );
1229 $text = preg_replace( "/^$h(.+)$h\\s*$/m",
1230 "<h$i>\\1</h$i>", $text );
1232 wfProfileOut( __METHOD__
);
1237 * Replace single quotes with HTML markup
1239 * @return string the altered text
1241 function doAllQuotes( $text ) {
1242 wfProfileIn( __METHOD__
);
1244 $lines = StringUtils
::explode( "\n", $text );
1245 foreach ( $lines as $line ) {
1246 $outtext .= $this->doQuotes( $line ) . "\n";
1248 $outtext = substr( $outtext, 0,-1 );
1249 wfProfileOut( __METHOD__
);
1254 * Helper function for doAllQuotes()
1256 public function doQuotes( $text ) {
1257 $arr = preg_split( "/(''+)/", $text, -1, PREG_SPLIT_DELIM_CAPTURE
);
1258 if ( count( $arr ) == 1 ) {
1261 # First, do some preliminary work. This may shift some apostrophes from
1262 # being mark-up to being text. It also counts the number of occurrences
1263 # of bold and italics mark-ups.
1267 foreach ( $arr as $r ) {
1268 if ( ( $i %
2 ) == 1 ) {
1269 # If there are ever four apostrophes, assume the first is supposed to
1270 # be text, and the remaining three constitute mark-up for bold text.
1271 if ( strlen( $arr[$i] ) == 4 ) {
1274 } elseif ( strlen( $arr[$i] ) > 5 ) {
1275 # If there are more than 5 apostrophes in a row, assume they're all
1276 # text except for the last 5.
1277 $arr[$i-1] .= str_repeat( "'", strlen( $arr[$i] ) - 5 );
1280 # Count the number of occurrences of bold and italics mark-ups.
1281 # We are not counting sequences of five apostrophes.
1282 if ( strlen( $arr[$i] ) == 2 ) {
1284 } elseif ( strlen( $arr[$i] ) == 3 ) {
1286 } elseif ( strlen( $arr[$i] ) == 5 ) {
1294 # If there is an odd number of both bold and italics, it is likely
1295 # that one of the bold ones was meant to be an apostrophe followed
1296 # by italics. Which one we cannot know for certain, but it is more
1297 # likely to be one that has a single-letter word before it.
1298 if ( ( $numbold %
2 == 1 ) && ( $numitalics %
2 == 1 ) ) {
1300 $firstsingleletterword = -1;
1301 $firstmultiletterword = -1;
1303 foreach ( $arr as $r ) {
1304 if ( ( $i %
2 == 1 ) and ( strlen( $r ) == 3 ) ) {
1305 $x1 = substr( $arr[$i-1], -1 );
1306 $x2 = substr( $arr[$i-1], -2, 1 );
1307 if ( $x1 === ' ' ) {
1308 if ( $firstspace == -1 ) {
1311 } elseif ( $x2 === ' ') {
1312 if ( $firstsingleletterword == -1 ) {
1313 $firstsingleletterword = $i;
1316 if ( $firstmultiletterword == -1 ) {
1317 $firstmultiletterword = $i;
1324 # If there is a single-letter word, use it!
1325 if ( $firstsingleletterword > -1 ) {
1326 $arr[$firstsingleletterword] = "''";
1327 $arr[$firstsingleletterword-1] .= "'";
1328 } elseif ( $firstmultiletterword > -1 ) {
1329 # If not, but there's a multi-letter word, use that one.
1330 $arr[$firstmultiletterword] = "''";
1331 $arr[$firstmultiletterword-1] .= "'";
1332 } elseif ( $firstspace > -1 ) {
1333 # ... otherwise use the first one that has neither.
1334 # (notice that it is possible for all three to be -1 if, for example,
1335 # there is only one pentuple-apostrophe in the line)
1336 $arr[$firstspace] = "''";
1337 $arr[$firstspace-1] .= "'";
1341 # Now let's actually convert our apostrophic mush to HTML!
1346 foreach ( $arr as $r ) {
1347 if ( ( $i %
2 ) == 0 ) {
1348 if ( $state === 'both' ) {
1354 if ( strlen( $r ) == 2 ) {
1355 if ( $state === 'i' ) {
1356 $output .= '</i>'; $state = '';
1357 } elseif ( $state === 'bi' ) {
1358 $output .= '</i>'; $state = 'b';
1359 } elseif ( $state === 'ib' ) {
1360 $output .= '</b></i><b>'; $state = 'b';
1361 } elseif ( $state === 'both' ) {
1362 $output .= '<b><i>'.$buffer.'</i>'; $state = 'b';
1363 } else { # $state can be 'b' or ''
1364 $output .= '<i>'; $state .= 'i';
1366 } elseif ( strlen( $r ) == 3 ) {
1367 if ( $state === 'b' ) {
1368 $output .= '</b>'; $state = '';
1369 } elseif ( $state === 'bi' ) {
1370 $output .= '</i></b><i>'; $state = 'i';
1371 } elseif ( $state === 'ib' ) {
1372 $output .= '</b>'; $state = 'i';
1373 } elseif ( $state === 'both' ) {
1374 $output .= '<i><b>'.$buffer.'</b>'; $state = 'i';
1375 } else { # $state can be 'i' or ''
1376 $output .= '<b>'; $state .= 'b';
1378 } elseif ( strlen( $r ) == 5 ) {
1379 if ( $state === 'b' ) {
1380 $output .= '</b><i>'; $state = 'i';
1381 } elseif ( $state === 'i' ) {
1382 $output .= '</i><b>'; $state = 'b';
1383 } elseif ( $state === 'bi' ) {
1384 $output .= '</i></b>'; $state = '';
1385 } elseif ( $state === 'ib' ) {
1386 $output .= '</b></i>'; $state = '';
1387 } elseif ( $state === 'both' ) {
1388 $output .= '<i><b>'.$buffer.'</b></i>'; $state = '';
1389 } else { # ($state == '')
1390 $buffer = ''; $state = 'both';
1396 # Now close all remaining tags. Notice that the order is important.
1397 if ( $state === 'b' ||
$state === 'ib' ) {
1400 if ( $state === 'i' ||
$state === 'bi' ||
$state === 'ib' ) {
1403 if ( $state === 'bi' ) {
1406 # There might be lonely ''''', so make sure we have a buffer
1407 if ( $state === 'both' && $buffer ) {
1408 $output .= '<b><i>'.$buffer.'</i></b>';
1415 * Replace external links (REL)
1417 * Note: this is all very hackish and the order of execution matters a lot.
1418 * Make sure to run maintenance/parserTests.php if you change this code.
1422 function replaceExternalLinks( $text ) {
1424 wfProfileIn( __METHOD__
);
1426 $sk = $this->mOptions
->getSkin( $this->mTitle
);
1428 $bits = preg_split( $this->mExtLinkBracketedRegex
, $text, -1, PREG_SPLIT_DELIM_CAPTURE
);
1429 $s = array_shift( $bits );
1432 while ( $i<count( $bits ) ) {
1434 $protocol = $bits[$i++
];
1435 $text = $bits[$i++
];
1436 $trail = $bits[$i++
];
1438 # The characters '<' and '>' (which were escaped by
1439 # removeHTMLtags()) should not be included in
1440 # URLs, per RFC 2396.
1442 if ( preg_match( '/&(lt|gt);/', $url, $m2, PREG_OFFSET_CAPTURE
) ) {
1443 $text = substr( $url, $m2[0][1] ) . ' ' . $text;
1444 $url = substr( $url, 0, $m2[0][1] );
1447 # If the link text is an image URL, replace it with an <img> tag
1448 # This happened by accident in the original parser, but some people used it extensively
1449 $img = $this->maybeMakeExternalImage( $text );
1450 if ( $img !== false ) {
1456 # Set linktype for CSS - if URL==text, link is essentially free
1457 $linktype = ( $text === $url ) ?
'free' : 'text';
1459 # No link text, e.g. [http://domain.tld/some.link]
1460 if ( $text == '' ) {
1461 # Autonumber if allowed. See bug #5918
1462 if ( strpos( wfUrlProtocols(), substr( $protocol, 0, strpos( $protocol, ':' ) ) ) !== false ) {
1463 $langObj = $this->getFunctionLang();
1464 $text = '[' . $langObj->formatNum( ++
$this->mAutonumber
) . ']';
1465 $linktype = 'autonumber';
1467 # Otherwise just use the URL
1468 $text = htmlspecialchars( $url );
1472 # Have link text, e.g. [http://domain.tld/some.link text]s
1474 list( $dtrail, $trail ) = Linker
::splitTrail( $trail );
1477 $text = $wgContLang->markNoConversion( $text );
1479 $url = Sanitizer
::cleanUrl( $url );
1481 # Use the encoded URL
1482 # This means that users can paste URLs directly into the text
1483 # Funny characters like ö aren't valid in URLs anyway
1484 # This was changed in August 2004
1485 $s .= $sk->makeExternalLink( $url, $text, false, $linktype,
1486 $this->getExternalLinkAttribs( $url ) ) . $dtrail . $trail;
1488 # Register link in the output object.
1489 # Replace unnecessary URL escape codes with the referenced character
1490 # This prevents spammers from hiding links from the filters
1491 $pasteurized = self
::replaceUnusualEscapes( $url );
1492 $this->mOutput
->addExternalLink( $pasteurized );
1495 wfProfileOut( __METHOD__
);
1500 * Get an associative array of additional HTML attributes appropriate for a
1501 * particular external link. This currently may include rel => nofollow
1502 * (depending on configuration, namespace, and the URL's domain) and/or a
1503 * target attribute (depending on configuration).
1505 * @param $url String: optional URL, to extract the domain from for rel =>
1506 * nofollow if appropriate
1507 * @return Array: associative array of HTML attributes
1509 function getExternalLinkAttribs( $url = false ) {
1511 global $wgNoFollowLinks, $wgNoFollowNsExceptions;
1512 $ns = $this->mTitle
->getNamespace();
1513 if ( $wgNoFollowLinks && !in_array( $ns, $wgNoFollowNsExceptions ) ) {
1514 $attribs['rel'] = 'nofollow';
1516 global $wgNoFollowDomainExceptions;
1517 if ( $wgNoFollowDomainExceptions ) {
1518 $bits = wfParseUrl( $url );
1519 if ( is_array( $bits ) && isset( $bits['host'] ) ) {
1520 foreach ( $wgNoFollowDomainExceptions as $domain ) {
1521 if ( substr( $bits['host'], -strlen( $domain ) ) == $domain ) {
1522 unset( $attribs['rel'] );
1529 if ( $this->mOptions
->getExternalLinkTarget() ) {
1530 $attribs['target'] = $this->mOptions
->getExternalLinkTarget();
1537 * Replace unusual URL escape codes with their equivalent characters
1539 * @param $url String
1542 * @todo This can merge genuinely required bits in the path or query string,
1543 * breaking legit URLs. A proper fix would treat the various parts of
1544 * the URL differently; as a workaround, just use the output for
1545 * statistical records, not for actual linking/output.
1547 static function replaceUnusualEscapes( $url ) {
1548 return preg_replace_callback( '/%[0-9A-Fa-f]{2}/',
1549 array( __CLASS__
, 'replaceUnusualEscapesCallback' ), $url );
1553 * Callback function used in replaceUnusualEscapes().
1554 * Replaces unusual URL escape codes with their equivalent character
1556 private static function replaceUnusualEscapesCallback( $matches ) {
1557 $char = urldecode( $matches[0] );
1558 $ord = ord( $char );
1559 # Is it an unsafe or HTTP reserved character according to RFC 1738?
1560 if ( $ord > 32 && $ord < 127 && strpos( '<>"#{}|\^~[]`;/?', $char ) === false ) {
1561 # No, shouldn't be escaped
1564 # Yes, leave it escaped
1570 * make an image if it's allowed, either through the global
1571 * option, through the exception, or through the on-wiki whitelist
1574 function maybeMakeExternalImage( $url ) {
1575 $sk = $this->mOptions
->getSkin( $this->mTitle
);
1576 $imagesfrom = $this->mOptions
->getAllowExternalImagesFrom();
1577 $imagesexception = !empty( $imagesfrom );
1579 # $imagesfrom could be either a single string or an array of strings, parse out the latter
1580 if ( $imagesexception && is_array( $imagesfrom ) ) {
1581 $imagematch = false;
1582 foreach ( $imagesfrom as $match ) {
1583 if ( strpos( $url, $match ) === 0 ) {
1588 } elseif ( $imagesexception ) {
1589 $imagematch = ( strpos( $url, $imagesfrom ) === 0 );
1591 $imagematch = false;
1593 if ( $this->mOptions
->getAllowExternalImages()
1594 ||
( $imagesexception && $imagematch ) ) {
1595 if ( preg_match( self
::EXT_IMAGE_REGEX
, $url ) ) {
1597 $text = $sk->makeExternalImage( $url );
1600 if ( !$text && $this->mOptions
->getEnableImageWhitelist()
1601 && preg_match( self
::EXT_IMAGE_REGEX
, $url ) ) {
1602 $whitelist = explode( "\n", wfMsgForContent( 'external_image_whitelist' ) );
1603 foreach ( $whitelist as $entry ) {
1604 # Sanitize the regex fragment, make it case-insensitive, ignore blank entries/comments
1605 if ( strpos( $entry, '#' ) === 0 ||
$entry === '' ) {
1608 if ( preg_match( '/' . str_replace( '/', '\\/', $entry ) . '/i', $url ) ) {
1609 # Image matches a whitelist entry
1610 $text = $sk->makeExternalImage( $url );
1619 * Process [[ ]] wikilinks
1620 * @return processed text
1624 function replaceInternalLinks( $s ) {
1625 $this->mLinkHolders
->merge( $this->replaceInternalLinks2( $s ) );
1630 * Process [[ ]] wikilinks (RIL)
1631 * @return LinkHolderArray
1635 function replaceInternalLinks2( &$s ) {
1638 wfProfileIn( __METHOD__
);
1640 wfProfileIn( __METHOD__
.'-setup' );
1641 static $tc = FALSE, $e1, $e1_img;
1642 # the % is needed to support urlencoded titles as well
1644 $tc = Title
::legalChars() . '#%';
1645 # Match a link having the form [[namespace:link|alternate]]trail
1646 $e1 = "/^([{$tc}]+)(?:\\|(.+?))?]](.*)\$/sD";
1647 # Match cases where there is no "]]", which might still be images
1648 $e1_img = "/^([{$tc}]+)\\|(.*)\$/sD";
1651 $sk = $this->mOptions
->getSkin( $this->mTitle
);
1652 $holders = new LinkHolderArray( $this );
1654 # split the entire text string on occurences of [[
1655 $a = StringUtils
::explode( '[[', ' ' . $s );
1656 # get the first element (all text up to first [[), and remove the space we added
1659 $line = $a->current(); # Workaround for broken ArrayIterator::next() that returns "void"
1660 $s = substr( $s, 1 );
1662 $useLinkPrefixExtension = $wgContLang->linkPrefixExtension();
1664 if ( $useLinkPrefixExtension ) {
1665 # Match the end of a line for a word that's not followed by whitespace,
1666 # e.g. in the case of 'The Arab al[[Razi]]', 'al' will be matched
1667 $e2 = wfMsgForContent( 'linkprefix' );
1670 if ( is_null( $this->mTitle
) ) {
1671 wfProfileOut( __METHOD__
.'-setup' );
1672 wfProfileOut( __METHOD__
);
1673 throw new MWException( __METHOD__
.": \$this->mTitle is null\n" );
1675 $nottalk = !$this->mTitle
->isTalkPage();
1677 if ( $useLinkPrefixExtension ) {
1679 if ( preg_match( $e2, $s, $m ) ) {
1680 $first_prefix = $m[2];
1682 $first_prefix = false;
1688 if ( $wgContLang->hasVariants() ) {
1689 $selflink = $wgContLang->autoConvertToAllVariants( $this->mTitle
->getPrefixedText() );
1691 $selflink = array( $this->mTitle
->getPrefixedText() );
1693 $useSubpages = $this->areSubpagesAllowed();
1694 wfProfileOut( __METHOD__
.'-setup' );
1696 # Loop for each link
1697 for ( ; $line !== false && $line !== null ; $a->next(), $line = $a->current() ) {
1698 # Check for excessive memory usage
1699 if ( $holders->isBig() ) {
1701 # Do the existence check, replace the link holders and clear the array
1702 $holders->replace( $s );
1706 if ( $useLinkPrefixExtension ) {
1707 wfProfileIn( __METHOD__
.'-prefixhandling' );
1708 if ( preg_match( $e2, $s, $m ) ) {
1715 if ( $first_prefix ) {
1716 $prefix = $first_prefix;
1717 $first_prefix = false;
1719 wfProfileOut( __METHOD__
.'-prefixhandling' );
1722 $might_be_img = false;
1724 wfProfileIn( __METHOD__
."-e1" );
1725 if ( preg_match( $e1, $line, $m ) ) { # page with normal text or alt
1727 # If we get a ] at the beginning of $m[3] that means we have a link that's something like:
1728 # [[Image:Foo.jpg|[http://example.com desc]]] <- having three ] in a row fucks up,
1729 # the real problem is with the $e1 regex
1732 # Still some problems for cases where the ] is meant to be outside punctuation,
1733 # and no image is in sight. See bug 2095.
1735 if ( $text !== '' &&
1736 substr( $m[3], 0, 1 ) === ']' &&
1737 strpos( $text, '[' ) !== false
1740 $text .= ']'; # so that replaceExternalLinks($text) works later
1741 $m[3] = substr( $m[3], 1 );
1743 # fix up urlencoded title texts
1744 if ( strpos( $m[1], '%' ) !== false ) {
1745 # Should anchors '#' also be rejected?
1746 $m[1] = str_replace( array('<', '>'), array('<', '>'), urldecode( $m[1] ) );
1749 } elseif ( preg_match( $e1_img, $line, $m ) ) { # Invalid, but might be an image with a link in its caption
1750 $might_be_img = true;
1752 if ( strpos( $m[1], '%' ) !== false ) {
1753 $m[1] = urldecode( $m[1] );
1756 } else { # Invalid form; output directly
1757 $s .= $prefix . '[[' . $line ;
1758 wfProfileOut( __METHOD__
."-e1" );
1761 wfProfileOut( __METHOD__
."-e1" );
1762 wfProfileIn( __METHOD__
."-misc" );
1764 # Don't allow internal links to pages containing
1765 # PROTO: where PROTO is a valid URL protocol; these
1766 # should be external links.
1767 if ( preg_match( '/^\b(?:' . wfUrlProtocols() . ')/', $m[1] ) ) {
1768 $s .= $prefix . '[[' . $line ;
1769 wfProfileOut( __METHOD__
."-misc" );
1773 # Make subpage if necessary
1774 if ( $useSubpages ) {
1775 $link = $this->maybeDoSubpageLink( $m[1], $text );
1780 $noforce = ( substr( $m[1], 0, 1 ) !== ':' );
1782 # Strip off leading ':'
1783 $link = substr( $link, 1 );
1786 wfProfileOut( __METHOD__
."-misc" );
1787 wfProfileIn( __METHOD__
."-title" );
1788 $nt = Title
::newFromText( $this->mStripState
->unstripNoWiki( $link ) );
1789 if ( $nt === null ) {
1790 $s .= $prefix . '[[' . $line;
1791 wfProfileOut( __METHOD__
."-title" );
1795 $ns = $nt->getNamespace();
1796 $iw = $nt->getInterWiki();
1797 wfProfileOut( __METHOD__
."-title" );
1799 if ( $might_be_img ) { # if this is actually an invalid link
1800 wfProfileIn( __METHOD__
."-might_be_img" );
1801 if ( $ns == NS_FILE
&& $noforce ) { # but might be an image
1804 # look at the next 'line' to see if we can close it there
1806 $next_line = $a->current();
1807 if ( $next_line === false ||
$next_line === null ) {
1810 $m = explode( ']]', $next_line, 3 );
1811 if ( count( $m ) == 3 ) {
1812 # the first ]] closes the inner link, the second the image
1814 $text .= "[[{$m[0]}]]{$m[1]}";
1817 } elseif ( count( $m ) == 2 ) {
1818 # if there's exactly one ]] that's fine, we'll keep looking
1819 $text .= "[[{$m[0]}]]{$m[1]}";
1821 # if $next_line is invalid too, we need look no further
1822 $text .= '[[' . $next_line;
1827 # we couldn't find the end of this imageLink, so output it raw
1828 # but don't ignore what might be perfectly normal links in the text we've examined
1829 $holders->merge( $this->replaceInternalLinks2( $text ) );
1830 $s .= "{$prefix}[[$link|$text";
1831 # note: no $trail, because without an end, there *is* no trail
1832 wfProfileOut( __METHOD__
."-might_be_img" );
1835 } else { # it's not an image, so output it raw
1836 $s .= "{$prefix}[[$link|$text";
1837 # note: no $trail, because without an end, there *is* no trail
1838 wfProfileOut( __METHOD__
."-might_be_img" );
1841 wfProfileOut( __METHOD__
."-might_be_img" );
1844 $wasblank = ( $text == '' );
1848 # Bug 4598 madness. Handle the quotes only if they come from the alternate part
1849 # [[Lista d''e paise d''o munno]] -> <a href="">Lista d''e paise d''o munno</a>
1850 # [[Criticism of Harry Potter|Criticism of ''Harry Potter'']] -> <a href="Criticism of Harry Potter">Criticism of <i>Harry Potter</i></a>
1851 $text = $this->doQuotes($text);
1854 # Link not escaped by : , create the various objects
1858 wfProfileIn( __METHOD__
."-interwiki" );
1859 if ( $iw && $this->mOptions
->getInterwikiMagic() && $nottalk && $wgContLang->getLanguageName( $iw ) ) {
1860 $this->mOutput
->addLanguageLink( $nt->getFullText() );
1861 $s = rtrim( $s . $prefix );
1862 $s .= trim( $trail, "\n" ) == '' ?
'': $prefix . $trail;
1863 wfProfileOut( __METHOD__
."-interwiki" );
1866 wfProfileOut( __METHOD__
."-interwiki" );
1868 if ( $ns == NS_FILE
) {
1869 wfProfileIn( __METHOD__
."-image" );
1870 if ( !wfIsBadImage( $nt->getDBkey(), $this->mTitle
) ) {
1872 # if no parameters were passed, $text
1873 # becomes something like "File:Foo.png",
1874 # which we don't want to pass on to the
1878 # recursively parse links inside the image caption
1879 # actually, this will parse them in any other parameters, too,
1880 # but it might be hard to fix that, and it doesn't matter ATM
1881 $text = $this->replaceExternalLinks( $text );
1882 $holders->merge( $this->replaceInternalLinks2( $text ) );
1884 # cloak any absolute URLs inside the image markup, so replaceExternalLinks() won't touch them
1885 $s .= $prefix . $this->armorLinks( $this->makeImage( $nt, $text, $holders ) ) . $trail;
1887 $s .= $prefix . $trail;
1889 $this->mOutput
->addImage( $nt->getDBkey() );
1890 wfProfileOut( __METHOD__
."-image" );
1895 if ( $ns == NS_CATEGORY
) {
1896 wfProfileIn( __METHOD__
."-category" );
1897 $s = rtrim( $s . "\n" ); # bug 87
1900 $sortkey = $this->getDefaultSort();
1904 $sortkey = Sanitizer
::decodeCharReferences( $sortkey );
1905 $sortkey = str_replace( "\n", '', $sortkey );
1906 $sortkey = $wgContLang->convertCategoryKey( $sortkey );
1907 $this->mOutput
->addCategory( $nt->getDBkey(), $sortkey );
1910 * Strip the whitespace Category links produce, see bug 87
1911 * @todo We might want to use trim($tmp, "\n") here.
1913 $s .= trim( $prefix . $trail, "\n" ) == '' ?
'': $prefix . $trail;
1915 wfProfileOut( __METHOD__
."-category" );
1920 # Self-link checking
1921 if ( $nt->getFragment() === '' && $ns != NS_SPECIAL
) {
1922 if ( in_array( $nt->getPrefixedText(), $selflink, true ) ) {
1923 $s .= $prefix . $sk->makeSelfLinkObj( $nt, $text, '', $trail );
1928 # NS_MEDIA is a pseudo-namespace for linking directly to a file
1929 # FIXME: Should do batch file existence checks, see comment below
1930 if ( $ns == NS_MEDIA
) {
1931 wfProfileIn( __METHOD__
."-media" );
1932 # Give extensions a chance to select the file revision for us
1933 $skip = $time = false;
1934 wfRunHooks( 'BeforeParserMakeImageLinkObj', array( &$this, &$nt, &$skip, &$time ) );
1936 $link = $sk->link( $nt );
1938 $link = $sk->makeMediaLinkObj( $nt, $text, $time );
1940 # Cloak with NOPARSE to avoid replacement in replaceExternalLinks
1941 $s .= $prefix . $this->armorLinks( $link ) . $trail;
1942 $this->mOutput
->addImage( $nt->getDBkey() );
1943 wfProfileOut( __METHOD__
."-media" );
1947 wfProfileIn( __METHOD__
."-always_known" );
1948 # Some titles, such as valid special pages or files in foreign repos, should
1949 # be shown as bluelinks even though they're not included in the page table
1951 # FIXME: isAlwaysKnown() can be expensive for file links; we should really do
1952 # batch file existence checks for NS_FILE and NS_MEDIA
1953 if ( $iw == '' && $nt->isAlwaysKnown() ) {
1954 $this->mOutput
->addLink( $nt );
1955 $s .= $this->makeKnownLinkHolder( $nt, $text, '', $trail, $prefix );
1957 # Links will be added to the output link list after checking
1958 $s .= $holders->makeHolder( $nt, $text, '', $trail, $prefix );
1960 wfProfileOut( __METHOD__
."-always_known" );
1962 wfProfileOut( __METHOD__
);
1967 * Make a link placeholder. The text returned can be later resolved to a real link with
1968 * replaceLinkHolders(). This is done for two reasons: firstly to avoid further
1969 * parsing of interwiki links, and secondly to allow all existence checks and
1970 * article length checks (for stub links) to be bundled into a single query.
1974 function makeLinkHolder( &$nt, $text = '', $query = '', $trail = '', $prefix = '' ) {
1975 return $this->mLinkHolders
->makeHolder( $nt, $text, $query, $trail, $prefix );
1979 * Render a forced-blue link inline; protect against double expansion of
1980 * URLs if we're in a mode that prepends full URL prefixes to internal links.
1981 * Since this little disaster has to split off the trail text to avoid
1982 * breaking URLs in the following text without breaking trails on the
1983 * wiki links, it's been made into a horrible function.
1986 * @param $text String
1987 * @param $query String
1988 * @param $trail String
1989 * @param $prefix String
1990 * @return String: HTML-wikitext mix oh yuck
1992 function makeKnownLinkHolder( $nt, $text = '', $query = '', $trail = '', $prefix = '' ) {
1993 list( $inside, $trail ) = Linker
::splitTrail( $trail );
1994 $sk = $this->mOptions
->getSkin( $this->mTitle
);
1995 # FIXME: use link() instead of deprecated makeKnownLinkObj()
1996 $link = $sk->makeKnownLinkObj( $nt, $text, $query, $inside, $prefix );
1997 return $this->armorLinks( $link ) . $trail;
2001 * Insert a NOPARSE hacky thing into any inline links in a chunk that's
2002 * going to go through further parsing steps before inline URL expansion.
2004 * Not needed quite as much as it used to be since free links are a bit
2005 * more sensible these days. But bracketed links are still an issue.
2007 * @param $text String: more-or-less HTML
2008 * @return String: less-or-more HTML with NOPARSE bits
2010 function armorLinks( $text ) {
2011 return preg_replace( '/\b(' . wfUrlProtocols() . ')/',
2012 "{$this->mUniqPrefix}NOPARSE$1", $text );
2016 * Return true if subpage links should be expanded on this page.
2019 function areSubpagesAllowed() {
2020 # Some namespaces don't allow subpages
2021 return MWNamespace
::hasSubpages( $this->mTitle
->getNamespace() );
2025 * Handle link to subpage if necessary
2027 * @param $target String: the source of the link
2028 * @param &$text String: the link text, modified as necessary
2029 * @return string the full name of the link
2032 function maybeDoSubpageLink( $target, &$text ) {
2033 return Linker
::normalizeSubpageLink( $this->mTitle
, $target, $text );
2037 * Used by doBlockLevels()
2040 function closeParagraph() {
2042 if ( $this->mLastSection
!= '' ) {
2043 $result = '</' . $this->mLastSection
. ">\n";
2045 $this->mInPre
= false;
2046 $this->mLastSection
= '';
2051 * getCommon() returns the length of the longest common substring
2052 * of both arguments, starting at the beginning of both.
2055 function getCommon( $st1, $st2 ) {
2056 $fl = strlen( $st1 );
2057 $shorter = strlen( $st2 );
2058 if ( $fl < $shorter ) {
2062 for ( $i = 0; $i < $shorter; ++
$i ) {
2063 if ( $st1{$i} != $st2{$i} ) {
2071 * These next three functions open, continue, and close the list
2072 * element appropriate to the prefix character passed into them.
2075 function openList( $char ) {
2076 $result = $this->closeParagraph();
2078 if ( '*' === $char ) {
2079 $result .= '<ul><li>';
2080 } elseif ( '#' === $char ) {
2081 $result .= '<ol><li>';
2082 } elseif ( ':' === $char ) {
2083 $result .= '<dl><dd>';
2084 } elseif ( ';' === $char ) {
2085 $result .= '<dl><dt>';
2086 $this->mDTopen
= true;
2088 $result = '<!-- ERR 1 -->';
2096 * @param $char String
2099 function nextItem( $char ) {
2100 if ( '*' === $char ||
'#' === $char ) {
2102 } elseif ( ':' === $char ||
';' === $char ) {
2104 if ( $this->mDTopen
) {
2107 if ( ';' === $char ) {
2108 $this->mDTopen
= true;
2109 return $close . '<dt>';
2111 $this->mDTopen
= false;
2112 return $close . '<dd>';
2115 return '<!-- ERR 2 -->';
2120 * @param $char String
2123 function closeList( $char ) {
2124 if ( '*' === $char ) {
2125 $text = '</li></ul>';
2126 } elseif ( '#' === $char ) {
2127 $text = '</li></ol>';
2128 } elseif ( ':' === $char ) {
2129 if ( $this->mDTopen
) {
2130 $this->mDTopen
= false;
2131 $text = '</dt></dl>';
2133 $text = '</dd></dl>';
2136 return '<!-- ERR 3 -->';
2143 * Make lists from lines starting with ':', '*', '#', etc. (DBL)
2145 * @param $text String
2146 * @param $linestart Boolean: whether or not this is at the start of a line.
2148 * @return string the lists rendered as HTML
2150 function doBlockLevels( $text, $linestart ) {
2151 wfProfileIn( __METHOD__
);
2153 # Parsing through the text line by line. The main thing
2154 # happening here is handling of block-level elements p, pre,
2155 # and making lists from lines starting with * # : etc.
2157 $textLines = StringUtils
::explode( "\n", $text );
2159 $lastPrefix = $output = '';
2160 $this->mDTopen
= $inBlockElem = false;
2162 $paragraphStack = false;
2164 foreach ( $textLines as $oLine ) {
2166 if ( !$linestart ) {
2176 $lastPrefixLength = strlen( $lastPrefix );
2177 $preCloseMatch = preg_match( '/<\\/pre/i', $oLine );
2178 $preOpenMatch = preg_match( '/<pre/i', $oLine );
2179 # If not in a <pre> element, scan for and figure out what prefixes are there.
2180 if ( !$this->mInPre
) {
2181 # Multiple prefixes may abut each other for nested lists.
2182 $prefixLength = strspn( $oLine, '*#:;' );
2183 $prefix = substr( $oLine, 0, $prefixLength );
2186 # ; and : are both from definition-lists, so they're equivalent
2187 # for the purposes of determining whether or not we need to open/close
2189 $prefix2 = str_replace( ';', ':', $prefix );
2190 $t = substr( $oLine, $prefixLength );
2191 $this->mInPre
= (bool)$preOpenMatch;
2193 # Don't interpret any other prefixes in preformatted text
2195 $prefix = $prefix2 = '';
2200 if ( $prefixLength && $lastPrefix === $prefix2 ) {
2201 # Same as the last item, so no need to deal with nesting or opening stuff
2202 $output .= $this->nextItem( substr( $prefix, -1 ) );
2203 $paragraphStack = false;
2205 if ( substr( $prefix, -1 ) === ';') {
2206 # The one nasty exception: definition lists work like this:
2207 # ; title : definition text
2208 # So we check for : in the remainder text to split up the
2209 # title and definition, without b0rking links.
2211 if ( $this->findColonNoLinks( $t, $term, $t2 ) !== false ) {
2213 $output .= $term . $this->nextItem( ':' );
2216 } elseif ( $prefixLength ||
$lastPrefixLength ) {
2217 # We need to open or close prefixes, or both.
2219 # Either open or close a level...
2220 $commonPrefixLength = $this->getCommon( $prefix, $lastPrefix );
2221 $paragraphStack = false;
2223 # Close all the prefixes which aren't shared.
2224 while ( $commonPrefixLength < $lastPrefixLength ) {
2225 $output .= $this->closeList( $lastPrefix[$lastPrefixLength-1] );
2226 --$lastPrefixLength;
2229 # Continue the current prefix if appropriate.
2230 if ( $prefixLength <= $commonPrefixLength && $commonPrefixLength > 0 ) {
2231 $output .= $this->nextItem( $prefix[$commonPrefixLength-1] );
2234 # Open prefixes where appropriate.
2235 while ( $prefixLength > $commonPrefixLength ) {
2236 $char = substr( $prefix, $commonPrefixLength, 1 );
2237 $output .= $this->openList( $char );
2239 if ( ';' === $char ) {
2240 # FIXME: This is dupe of code above
2241 if ( $this->findColonNoLinks( $t, $term, $t2 ) !== false ) {
2243 $output .= $term . $this->nextItem( ':' );
2246 ++
$commonPrefixLength;
2248 $lastPrefix = $prefix2;
2251 # If we have no prefixes, go to paragraph mode.
2252 if ( 0 == $prefixLength ) {
2253 wfProfileIn( __METHOD__
."-paragraph" );
2254 # No prefix (not in list)--go to paragraph mode
2255 # XXX: use a stack for nestable elements like span, table and div
2256 $openmatch = preg_match('/(?:<table|<blockquote|<h1|<h2|<h3|<h4|<h5|<h6|<pre|<tr|<p|<ul|<ol|<li|<\\/tr|<\\/td|<\\/th)/iS', $t );
2257 $closematch = preg_match(
2258 '/(?:<\\/table|<\\/blockquote|<\\/h1|<\\/h2|<\\/h3|<\\/h4|<\\/h5|<\\/h6|'.
2259 '<td|<th|<\\/?div|<hr|<\\/pre|<\\/p|'.$this->mUniqPrefix
.'-pre|<\\/li|<\\/ul|<\\/ol|<\\/?center)/iS', $t );
2260 if ( $openmatch or $closematch ) {
2261 $paragraphStack = false;
2262 #Â TODO bug 5718: paragraph closed
2263 $output .= $this->closeParagraph();
2264 if ( $preOpenMatch and !$preCloseMatch ) {
2265 $this->mInPre
= true;
2267 $inBlockElem = !$closematch;
2268 } elseif ( !$inBlockElem && !$this->mInPre
) {
2269 if ( ' ' == substr( $t, 0, 1 ) and ( $this->mLastSection
=== 'pre' ||
trim( $t ) != '' ) ) {
2271 if ( $this->mLastSection
!== 'pre' ) {
2272 $paragraphStack = false;
2273 $output .= $this->closeParagraph().'<pre>';
2274 $this->mLastSection
= 'pre';
2276 $t = substr( $t, 1 );
2279 if ( trim( $t ) === '' ) {
2280 if ( $paragraphStack ) {
2281 $output .= $paragraphStack.'<br />';
2282 $paragraphStack = false;
2283 $this->mLastSection
= 'p';
2285 if ( $this->mLastSection
!== 'p' ) {
2286 $output .= $this->closeParagraph();
2287 $this->mLastSection
= '';
2288 $paragraphStack = '<p>';
2290 $paragraphStack = '</p><p>';
2294 if ( $paragraphStack ) {
2295 $output .= $paragraphStack;
2296 $paragraphStack = false;
2297 $this->mLastSection
= 'p';
2298 } elseif ( $this->mLastSection
!== 'p' ) {
2299 $output .= $this->closeParagraph().'<p>';
2300 $this->mLastSection
= 'p';
2305 wfProfileOut( __METHOD__
."-paragraph" );
2307 # somewhere above we forget to get out of pre block (bug 785)
2308 if ( $preCloseMatch && $this->mInPre
) {
2309 $this->mInPre
= false;
2311 if ( $paragraphStack === false ) {
2315 while ( $prefixLength ) {
2316 $output .= $this->closeList( $prefix2[$prefixLength-1] );
2319 if ( $this->mLastSection
!= '' ) {
2320 $output .= '</' . $this->mLastSection
. '>';
2321 $this->mLastSection
= '';
2324 wfProfileOut( __METHOD__
);
2329 * Split up a string on ':', ignoring any occurences inside tags
2330 * to prevent illegal overlapping.
2332 * @param $str String: the string to split
2333 * @param &$before String: set to everything before the ':'
2334 * @param &$after String: set to everything after the ':'
2335 * return String: the position of the ':', or false if none found
2337 function findColonNoLinks( $str, &$before, &$after ) {
2338 wfProfileIn( __METHOD__
);
2340 $pos = strpos( $str, ':' );
2341 if ( $pos === false ) {
2343 wfProfileOut( __METHOD__
);
2347 $lt = strpos( $str, '<' );
2348 if ( $lt === false ||
$lt > $pos ) {
2349 # Easy; no tag nesting to worry about
2350 $before = substr( $str, 0, $pos );
2351 $after = substr( $str, $pos+
1 );
2352 wfProfileOut( __METHOD__
);
2356 # Ugly state machine to walk through avoiding tags.
2357 $state = self
::COLON_STATE_TEXT
;
2359 $len = strlen( $str );
2360 for( $i = 0; $i < $len; $i++
) {
2364 # (Using the number is a performance hack for common cases)
2365 case 0: # self::COLON_STATE_TEXT:
2368 # Could be either a <start> tag or an </end> tag
2369 $state = self
::COLON_STATE_TAGSTART
;
2372 if ( $stack == 0 ) {
2374 $before = substr( $str, 0, $i );
2375 $after = substr( $str, $i +
1 );
2376 wfProfileOut( __METHOD__
);
2379 # Embedded in a tag; don't break it.
2382 # Skip ahead looking for something interesting
2383 $colon = strpos( $str, ':', $i );
2384 if ( $colon === false ) {
2385 # Nothing else interesting
2386 wfProfileOut( __METHOD__
);
2389 $lt = strpos( $str, '<', $i );
2390 if ( $stack === 0 ) {
2391 if ( $lt === false ||
$colon < $lt ) {
2393 $before = substr( $str, 0, $colon );
2394 $after = substr( $str, $colon +
1 );
2395 wfProfileOut( __METHOD__
);
2399 if ( $lt === false ) {
2400 # Nothing else interesting to find; abort!
2401 # We're nested, but there's no close tags left. Abort!
2404 # Skip ahead to next tag start
2406 $state = self
::COLON_STATE_TAGSTART
;
2409 case 1: # self::COLON_STATE_TAG:
2414 $state = self
::COLON_STATE_TEXT
;
2417 # Slash may be followed by >?
2418 $state = self
::COLON_STATE_TAGSLASH
;
2424 case 2: # self::COLON_STATE_TAGSTART:
2427 $state = self
::COLON_STATE_CLOSETAG
;
2430 $state = self
::COLON_STATE_COMMENT
;
2433 # Illegal early close? This shouldn't happen D:
2434 $state = self
::COLON_STATE_TEXT
;
2437 $state = self
::COLON_STATE_TAG
;
2440 case 3: # self::COLON_STATE_CLOSETAG:
2445 wfDebug( __METHOD__
.": Invalid input; too many close tags\n" );
2446 wfProfileOut( __METHOD__
);
2449 $state = self
::COLON_STATE_TEXT
;
2452 case self
::COLON_STATE_TAGSLASH
:
2454 # Yes, a self-closed tag <blah/>
2455 $state = self
::COLON_STATE_TEXT
;
2457 # Probably we're jumping the gun, and this is an attribute
2458 $state = self
::COLON_STATE_TAG
;
2461 case 5: # self::COLON_STATE_COMMENT:
2463 $state = self
::COLON_STATE_COMMENTDASH
;
2466 case self
::COLON_STATE_COMMENTDASH
:
2468 $state = self
::COLON_STATE_COMMENTDASHDASH
;
2470 $state = self
::COLON_STATE_COMMENT
;
2473 case self
::COLON_STATE_COMMENTDASHDASH
:
2475 $state = self
::COLON_STATE_TEXT
;
2477 $state = self
::COLON_STATE_COMMENT
;
2481 throw new MWException( "State machine error in " . __METHOD__
);
2485 wfDebug( __METHOD__
.": Invalid input; not enough close tags (stack $stack, state $state)\n" );
2488 wfProfileOut( __METHOD__
);
2493 * Return value of a magic variable (like PAGENAME)
2497 function getVariableValue( $index, $frame=false ) {
2498 global $wgContLang, $wgSitename, $wgServer, $wgServerName;
2499 global $wgArticlePath, $wgScriptPath, $wgStylePath;
2502 * Some of these require message or data lookups and can be
2503 * expensive to check many times.
2505 if ( wfRunHooks( 'ParserGetVariableValueVarCache', array( &$this, &$this->mVarCache
) ) ) {
2506 if ( isset( $this->mVarCache
[$index] ) ) {
2507 return $this->mVarCache
[$index];
2511 $ts = wfTimestamp( TS_UNIX
, $this->mOptions
->getTimestamp() );
2512 wfRunHooks( 'ParserGetVariableValueTs', array( &$this, &$ts ) );
2515 global $wgLocaltimezone;
2516 if ( isset( $wgLocaltimezone ) ) {
2517 $oldtz = date_default_timezone_get();
2518 date_default_timezone_set( $wgLocaltimezone );
2521 $localTimestamp = date( 'YmdHis', $ts );
2522 $localMonth = date( 'm', $ts );
2523 $localMonth1 = date( 'n', $ts );
2524 $localMonthName = date( 'n', $ts );
2525 $localDay = date( 'j', $ts );
2526 $localDay2 = date( 'd', $ts );
2527 $localDayOfWeek = date( 'w', $ts );
2528 $localWeek = date( 'W', $ts );
2529 $localYear = date( 'Y', $ts );
2530 $localHour = date( 'H', $ts );
2531 if ( isset( $wgLocaltimezone ) ) {
2532 date_default_timezone_set( $oldtz );
2536 case 'currentmonth':
2537 $value = $wgContLang->formatNum( gmdate( 'm', $ts ) );
2539 case 'currentmonth1':
2540 $value = $wgContLang->formatNum( gmdate( 'n', $ts ) );
2542 case 'currentmonthname':
2543 $value = $wgContLang->getMonthName( gmdate( 'n', $ts ) );
2545 case 'currentmonthnamegen':
2546 $value = $wgContLang->getMonthNameGen( gmdate( 'n', $ts ) );
2548 case 'currentmonthabbrev':
2549 $value = $wgContLang->getMonthAbbreviation( gmdate( 'n', $ts ) );
2552 $value = $wgContLang->formatNum( gmdate( 'j', $ts ) );
2555 $value = $wgContLang->formatNum( gmdate( 'd', $ts ) );
2558 $value = $wgContLang->formatNum( $localMonth );
2561 $value = $wgContLang->formatNum( $localMonth1 );
2563 case 'localmonthname':
2564 $value = $wgContLang->getMonthName( $localMonthName );
2566 case 'localmonthnamegen':
2567 $value = $wgContLang->getMonthNameGen( $localMonthName );
2569 case 'localmonthabbrev':
2570 $value = $wgContLang->getMonthAbbreviation( $localMonthName );
2573 $value = $wgContLang->formatNum( $localDay );
2576 $value = $wgContLang->formatNum( $localDay2 );
2579 $value = wfEscapeWikiText( $this->mTitle
->getText() );
2582 $value = $this->mTitle
->getPartialURL();
2584 case 'fullpagename':
2585 $value = wfEscapeWikiText( $this->mTitle
->getPrefixedText() );
2587 case 'fullpagenamee':
2588 $value = $this->mTitle
->getPrefixedURL();
2591 $value = wfEscapeWikiText( $this->mTitle
->getSubpageText() );
2593 case 'subpagenamee':
2594 $value = $this->mTitle
->getSubpageUrlForm();
2596 case 'basepagename':
2597 $value = wfEscapeWikiText( $this->mTitle
->getBaseText() );
2599 case 'basepagenamee':
2600 $value = wfUrlEncode( str_replace( ' ', '_', $this->mTitle
->getBaseText() ) );
2602 case 'talkpagename':
2603 if ( $this->mTitle
->canTalk() ) {
2604 $talkPage = $this->mTitle
->getTalkPage();
2605 $value = wfEscapeWikiText( $talkPage->getPrefixedText() );
2610 case 'talkpagenamee':
2611 if ( $this->mTitle
->canTalk() ) {
2612 $talkPage = $this->mTitle
->getTalkPage();
2613 $value = $talkPage->getPrefixedUrl();
2618 case 'subjectpagename':
2619 $subjPage = $this->mTitle
->getSubjectPage();
2620 $value = wfEscapeWikiText( $subjPage->getPrefixedText() );
2622 case 'subjectpagenamee':
2623 $subjPage = $this->mTitle
->getSubjectPage();
2624 $value = $subjPage->getPrefixedUrl();
2627 # Let the edit saving system know we should parse the page
2628 # *after* a revision ID has been assigned.
2629 $this->mOutput
->setFlag( 'vary-revision' );
2630 wfDebug( __METHOD__
. ": {{REVISIONID}} used, setting vary-revision...\n" );
2631 $value = $this->mRevisionId
;
2634 # Let the edit saving system know we should parse the page
2635 # *after* a revision ID has been assigned. This is for null edits.
2636 $this->mOutput
->setFlag( 'vary-revision' );
2637 wfDebug( __METHOD__
. ": {{REVISIONDAY}} used, setting vary-revision...\n" );
2638 $value = intval( substr( $this->getRevisionTimestamp(), 6, 2 ) );
2640 case 'revisionday2':
2641 # Let the edit saving system know we should parse the page
2642 # *after* a revision ID has been assigned. This is for null edits.
2643 $this->mOutput
->setFlag( 'vary-revision' );
2644 wfDebug( __METHOD__
. ": {{REVISIONDAY2}} used, setting vary-revision...\n" );
2645 $value = substr( $this->getRevisionTimestamp(), 6, 2 );
2647 case 'revisionmonth':
2648 # Let the edit saving system know we should parse the page
2649 # *after* a revision ID has been assigned. This is for null edits.
2650 $this->mOutput
->setFlag( 'vary-revision' );
2651 wfDebug( __METHOD__
. ": {{REVISIONMONTH}} used, setting vary-revision...\n" );
2652 $value = substr( $this->getRevisionTimestamp(), 4, 2 );
2654 case 'revisionmonth1':
2655 # Let the edit saving system know we should parse the page
2656 # *after* a revision ID has been assigned. This is for null edits.
2657 $this->mOutput
->setFlag( 'vary-revision' );
2658 wfDebug( __METHOD__
. ": {{REVISIONMONTH1}} used, setting vary-revision...\n" );
2659 $value = intval( substr( $this->getRevisionTimestamp(), 4, 2 ) );
2661 case 'revisionyear':
2662 # Let the edit saving system know we should parse the page
2663 # *after* a revision ID has been assigned. This is for null edits.
2664 $this->mOutput
->setFlag( 'vary-revision' );
2665 wfDebug( __METHOD__
. ": {{REVISIONYEAR}} used, setting vary-revision...\n" );
2666 $value = substr( $this->getRevisionTimestamp(), 0, 4 );
2668 case 'revisiontimestamp':
2669 # Let the edit saving system know we should parse the page
2670 # *after* a revision ID has been assigned. This is for null edits.
2671 $this->mOutput
->setFlag( 'vary-revision' );
2672 wfDebug( __METHOD__
. ": {{REVISIONTIMESTAMP}} used, setting vary-revision...\n" );
2673 $value = $this->getRevisionTimestamp();
2675 case 'revisionuser':
2676 # Let the edit saving system know we should parse the page
2677 # *after* a revision ID has been assigned. This is for null edits.
2678 $this->mOutput
->setFlag( 'vary-revision' );
2679 wfDebug( __METHOD__
. ": {{REVISIONUSER}} used, setting vary-revision...\n" );
2680 $value = $this->getRevisionUser();
2683 $value = str_replace( '_',' ',$wgContLang->getNsText( $this->mTitle
->getNamespace() ) );
2686 $value = wfUrlencode( $wgContLang->getNsText( $this->mTitle
->getNamespace() ) );
2689 $value = $this->mTitle
->canTalk() ?
str_replace( '_',' ',$this->mTitle
->getTalkNsText() ) : '';
2692 $value = $this->mTitle
->canTalk() ?
wfUrlencode( $this->mTitle
->getTalkNsText() ) : '';
2694 case 'subjectspace':
2695 $value = $this->mTitle
->getSubjectNsText();
2697 case 'subjectspacee':
2698 $value = ( wfUrlencode( $this->mTitle
->getSubjectNsText() ) );
2700 case 'currentdayname':
2701 $value = $wgContLang->getWeekdayName( gmdate( 'w', $ts ) +
1 );
2704 $value = $wgContLang->formatNum( gmdate( 'Y', $ts ), true );
2707 $value = $wgContLang->time( wfTimestamp( TS_MW
, $ts ), false, false );
2710 $value = $wgContLang->formatNum( gmdate( 'H', $ts ), true );
2713 # @bug 4594 PHP5 has it zero padded, PHP4 does not, cast to
2714 # int to remove the padding
2715 $value = $wgContLang->formatNum( (int)gmdate( 'W', $ts ) );
2718 $value = $wgContLang->formatNum( gmdate( 'w', $ts ) );
2720 case 'localdayname':
2721 $value = $wgContLang->getWeekdayName( $localDayOfWeek +
1 );
2724 $value = $wgContLang->formatNum( $localYear, true );
2727 $value = $wgContLang->time( $localTimestamp, false, false );
2730 $value = $wgContLang->formatNum( $localHour, true );
2733 # @bug 4594 PHP5 has it zero padded, PHP4 does not, cast to
2734 # int to remove the padding
2735 $value = $wgContLang->formatNum( (int)$localWeek );
2738 $value = $wgContLang->formatNum( $localDayOfWeek );
2740 case 'numberofarticles':
2741 $value = $wgContLang->formatNum( SiteStats
::articles() );
2743 case 'numberoffiles':
2744 $value = $wgContLang->formatNum( SiteStats
::images() );
2746 case 'numberofusers':
2747 $value = $wgContLang->formatNum( SiteStats
::users() );
2749 case 'numberofactiveusers':
2750 $value = $wgContLang->formatNum( SiteStats
::activeUsers() );
2752 case 'numberofpages':
2753 $value = $wgContLang->formatNum( SiteStats
::pages() );
2755 case 'numberofadmins':
2756 $value = $wgContLang->formatNum( SiteStats
::numberingroup( 'sysop' ) );
2758 case 'numberofedits':
2759 $value = $wgContLang->formatNum( SiteStats
::edits() );
2761 case 'numberofviews':
2762 $value = $wgContLang->formatNum( SiteStats
::views() );
2764 case 'currenttimestamp':
2765 $value = wfTimestamp( TS_MW
, $ts );
2767 case 'localtimestamp':
2768 $value = $localTimestamp;
2770 case 'currentversion':
2771 $value = SpecialVersion
::getVersion();
2774 return $wgArticlePath;
2780 return $wgServerName;
2782 return $wgScriptPath;
2784 return $wgStylePath;
2785 case 'directionmark':
2786 return $wgContLang->getDirMark();
2787 case 'contentlanguage':
2788 global $wgLanguageCode;
2789 return $wgLanguageCode;
2792 if ( wfRunHooks( 'ParserGetVariableValueSwitch', array( &$this, &$this->mVarCache
, &$index, &$ret, &$frame ) ) ) {
2800 $this->mVarCache
[$index] = $value;
2806 * initialise the magic variables (like CURRENTMONTHNAME) and substitution modifiers
2810 function initialiseVariables() {
2811 wfProfileIn( __METHOD__
);
2812 $variableIDs = MagicWord
::getVariableIDs();
2813 $substIDs = MagicWord
::getSubstIDs();
2815 $this->mVariables
= new MagicWordArray( $variableIDs );
2816 $this->mSubstWords
= new MagicWordArray( $substIDs );
2817 wfProfileOut( __METHOD__
);
2821 * Preprocess some wikitext and return the document tree.
2822 * This is the ghost of replace_variables().
2824 * @param $text String: The text to parse
2825 * @param $flags Integer: bitwise combination of:
2826 * self::PTD_FOR_INCLUSION Handle <noinclude>/<includeonly> as if the text is being
2827 * included. Default is to assume a direct page view.
2829 * The generated DOM tree must depend only on the input text and the flags.
2830 * The DOM tree must be the same in OT_HTML and OT_WIKI mode, to avoid a regression of bug 4899.
2832 * Any flag added to the $flags parameter here, or any other parameter liable to cause a
2833 * change in the DOM tree for a given text, must be passed through the section identifier
2834 * in the section edit link and thus back to extractSections().
2836 * The output of this function is currently only cached in process memory, but a persistent
2837 * cache may be implemented at a later date which takes further advantage of these strict
2838 * dependency requirements.
2842 function preprocessToDom( $text, $flags = 0 ) {
2843 $dom = $this->getPreprocessor()->preprocessToObj( $text, $flags );
2848 * Return a three-element array: leading whitespace, string contents, trailing whitespace
2850 public static function splitWhitespace( $s ) {
2851 $ltrimmed = ltrim( $s );
2852 $w1 = substr( $s, 0, strlen( $s ) - strlen( $ltrimmed ) );
2853 $trimmed = rtrim( $ltrimmed );
2854 $diff = strlen( $ltrimmed ) - strlen( $trimmed );
2856 $w2 = substr( $ltrimmed, -$diff );
2860 return array( $w1, $trimmed, $w2 );
2864 * Replace magic variables, templates, and template arguments
2865 * with the appropriate text. Templates are substituted recursively,
2866 * taking care to avoid infinite loops.
2868 * Note that the substitution depends on value of $mOutputType:
2869 * self::OT_WIKI: only {{subst:}} templates
2870 * self::OT_PREPROCESS: templates but not extension tags
2871 * self::OT_HTML: all templates and extension tags
2873 * @param $text String: the text to transform
2874 * @param $frame PPFrame Object describing the arguments passed to the template.
2875 * Arguments may also be provided as an associative array, as was the usual case before MW1.12.
2876 * Providing arguments this way may be useful for extensions wishing to perform variable replacement explicitly.
2877 * @param $argsOnly Boolean: only do argument (triple-brace) expansion, not double-brace expansion
2880 function replaceVariables( $text, $frame = false, $argsOnly = false ) {
2881 # Is there any text? Also, Prevent too big inclusions!
2882 if ( strlen( $text ) < 1 ||
strlen( $text ) > $this->mOptions
->getMaxIncludeSize() ) {
2885 wfProfileIn( __METHOD__
);
2887 if ( $frame === false ) {
2888 $frame = $this->getPreprocessor()->newFrame();
2889 } elseif ( !( $frame instanceof PPFrame
) ) {
2890 wfDebug( __METHOD__
." called using plain parameters instead of a PPFrame instance. Creating custom frame.\n" );
2891 $frame = $this->getPreprocessor()->newCustomFrame( $frame );
2894 $dom = $this->preprocessToDom( $text );
2895 $flags = $argsOnly ? PPFrame
::NO_TEMPLATES
: 0;
2896 $text = $frame->expand( $dom, $flags );
2898 wfProfileOut( __METHOD__
);
2902 # Clean up argument array - refactored in 1.9 so parserfunctions can use it, too.
2903 static function createAssocArgs( $args ) {
2904 $assocArgs = array();
2906 foreach ( $args as $arg ) {
2907 $eqpos = strpos( $arg, '=' );
2908 if ( $eqpos === false ) {
2909 $assocArgs[$index++
] = $arg;
2911 $name = trim( substr( $arg, 0, $eqpos ) );
2912 $value = trim( substr( $arg, $eqpos+
1 ) );
2913 if ( $value === false ) {
2916 if ( $name !== false ) {
2917 $assocArgs[$name] = $value;
2926 * Warn the user when a parser limitation is reached
2927 * Will warn at most once the user per limitation type
2929 * @param $limitationType String: should be one of:
2930 * 'expensive-parserfunction' (corresponding messages:
2931 * 'expensive-parserfunction-warning',
2932 * 'expensive-parserfunction-category')
2933 * 'post-expand-template-argument' (corresponding messages:
2934 * 'post-expand-template-argument-warning',
2935 * 'post-expand-template-argument-category')
2936 * 'post-expand-template-inclusion' (corresponding messages:
2937 * 'post-expand-template-inclusion-warning',
2938 * 'post-expand-template-inclusion-category')
2939 * @param $current Current value
2940 * @param $max Maximum allowed, when an explicit limit has been
2941 * exceeded, provide the values (optional)
2943 function limitationWarn( $limitationType, $current=null, $max=null) {
2944 # does no harm if $current and $max are present but are unnecessary for the message
2945 $warning = wfMsgExt( "$limitationType-warning", array( 'parsemag', 'escape' ), $current, $max );
2946 $this->mOutput
->addWarning( $warning );
2947 $this->addTrackingCategory( "$limitationType-category" );
2951 * Return the text of a template, after recursively
2952 * replacing any variables or templates within the template.
2954 * @param $piece Array: the parts of the template
2955 * $piece['title']: the title, i.e. the part before the |
2956 * $piece['parts']: the parameter array
2957 * $piece['lineStart']: whether the brace was at the start of a line
2958 * @param $frame PPFrame The current frame, contains template arguments
2959 * @return String: the text of the template
2962 function braceSubstitution( $piece, $frame ) {
2963 global $wgContLang, $wgNonincludableNamespaces;
2964 wfProfileIn( __METHOD__
);
2965 wfProfileIn( __METHOD__
.'-setup' );
2968 $found = false; # $text has been filled
2969 $nowiki = false; # wiki markup in $text should be escaped
2970 $isHTML = false; # $text is HTML, armour it against wikitext transformation
2971 $forceRawInterwiki = false; # Force interwiki transclusion to be done in raw mode not rendered
2972 $isChildObj = false; # $text is a DOM node needing expansion in a child frame
2973 $isLocalObj = false; # $text is a DOM node needing expansion in the current frame
2975 # Title object, where $text came from
2978 # $part1 is the bit before the first |, and must contain only title characters.
2979 # Various prefixes will be stripped from it later.
2980 $titleWithSpaces = $frame->expand( $piece['title'] );
2981 $part1 = trim( $titleWithSpaces );
2984 # Original title text preserved for various purposes
2985 $originalTitle = $part1;
2987 # $args is a list of argument nodes, starting from index 0, not including $part1
2988 $args = ( null == $piece['parts'] ) ?
array() : $piece['parts'];
2989 wfProfileOut( __METHOD__
.'-setup' );
2992 wfProfileIn( __METHOD__
.'-modifiers' );
2995 $substMatch = $this->mSubstWords
->matchStartAndRemove( $part1 );
2997 # Possibilities for substMatch: "subst", "safesubst" or FALSE
2998 # Decide whether to expand template or keep wikitext as-is.
2999 if ( $this->ot
['wiki'] ) {
3000 if ( $substMatch === false ) {
3001 $literal = true; # literal when in PST with no prefix
3003 $literal = false; # expand when in PST with subst: or safesubst:
3006 if ( $substMatch == 'subst' ) {
3007 $literal = true; # literal when not in PST with plain subst:
3009 $literal = false; # expand when not in PST with safesubst: or no prefix
3013 $text = $frame->virtualBracketedImplode( '{{', '|', '}}', $titleWithSpaces, $args );
3020 if ( !$found && $args->getLength() == 0 ) {
3021 $id = $this->mVariables
->matchStartToEnd( $part1 );
3022 if ( $id !== false ) {
3023 $text = $this->getVariableValue( $id, $frame );
3024 if ( MagicWord
::getCacheTTL( $id ) > -1 ) {
3025 $this->mOutput
->updateCacheExpiry( MagicWord
::getCacheTTL( $id ) );
3031 # MSG, MSGNW and RAW
3034 $mwMsgnw = MagicWord
::get( 'msgnw' );
3035 if ( $mwMsgnw->matchStartAndRemove( $part1 ) ) {
3038 # Remove obsolete MSG:
3039 $mwMsg = MagicWord
::get( 'msg' );
3040 $mwMsg->matchStartAndRemove( $part1 );
3044 $mwRaw = MagicWord
::get( 'raw' );
3045 if ( $mwRaw->matchStartAndRemove( $part1 ) ) {
3046 $forceRawInterwiki = true;
3049 wfProfileOut( __METHOD__
.'-modifiers' );
3053 wfProfileIn( __METHOD__
. '-pfunc' );
3055 $colonPos = strpos( $part1, ':' );
3056 if ( $colonPos !== false ) {
3057 # Case sensitive functions
3058 $function = substr( $part1, 0, $colonPos );
3059 if ( isset( $this->mFunctionSynonyms
[1][$function] ) ) {
3060 $function = $this->mFunctionSynonyms
[1][$function];
3062 # Case insensitive functions
3063 $function = $wgContLang->lc( $function );
3064 if ( isset( $this->mFunctionSynonyms
[0][$function] ) ) {
3065 $function = $this->mFunctionSynonyms
[0][$function];
3071 list( $callback, $flags ) = $this->mFunctionHooks
[$function];
3072 $initialArgs = array( &$this );
3073 $funcArgs = array( trim( substr( $part1, $colonPos +
1 ) ) );
3074 if ( $flags & SFH_OBJECT_ARGS
) {
3075 # Add a frame parameter, and pass the arguments as an array
3076 $allArgs = $initialArgs;
3077 $allArgs[] = $frame;
3078 for ( $i = 0; $i < $args->getLength(); $i++
) {
3079 $funcArgs[] = $args->item( $i );
3081 $allArgs[] = $funcArgs;
3083 # Convert arguments to plain text
3084 for ( $i = 0; $i < $args->getLength(); $i++
) {
3085 $funcArgs[] = trim( $frame->expand( $args->item( $i ) ) );
3087 $allArgs = array_merge( $initialArgs, $funcArgs );
3090 # Workaround for PHP bug 35229 and similar
3091 if ( !is_callable( $callback ) ) {
3092 wfProfileOut( __METHOD__
. '-pfunc' );
3093 wfProfileOut( __METHOD__
);
3094 throw new MWException( "Tag hook for $function is not callable\n" );
3096 $result = call_user_func_array( $callback, $allArgs );
3099 $preprocessFlags = 0;
3101 if ( is_array( $result ) ) {
3102 if ( isset( $result[0] ) ) {
3104 unset( $result[0] );
3107 # Extract flags into the local scope
3108 # This allows callers to set flags such as nowiki, found, etc.
3114 $text = $this->preprocessToDom( $text, $preprocessFlags );
3119 wfProfileOut( __METHOD__
. '-pfunc' );
3122 # Finish mangling title and then check for loops.
3123 # Set $title to a Title object and $titleText to the PDBK
3126 # Split the title into page and subpage
3128 $part1 = $this->maybeDoSubpageLink( $part1, $subpage );
3129 if ( $subpage !== '' ) {
3130 $ns = $this->mTitle
->getNamespace();
3132 $title = Title
::newFromText( $part1, $ns );
3134 $titleText = $title->getPrefixedText();
3135 # Check for language variants if the template is not found
3136 if ( $wgContLang->hasVariants() && $title->getArticleID() == 0 ) {
3137 $wgContLang->findVariantLink( $part1, $title, true );
3139 # Do recursion depth check
3140 $limit = $this->mOptions
->getMaxTemplateDepth();
3141 if ( $frame->depth
>= $limit ) {
3143 $text = '<span class="error">'
3144 . wfMsgForContent( 'parser-template-recursion-depth-warning', $limit )
3150 # Load from database
3151 if ( !$found && $title ) {
3152 wfProfileIn( __METHOD__
. '-loadtpl' );
3153 if ( !$title->isExternal() ) {
3154 if ( $title->getNamespace() == NS_SPECIAL
3155 && $this->mOptions
->getAllowSpecialInclusion()
3156 && $this->ot
['html'] )
3158 $text = SpecialPage
::capturePath( $title );
3159 if ( is_string( $text ) ) {
3162 $this->disableCache();
3164 } elseif ( $wgNonincludableNamespaces && in_array( $title->getNamespace(), $wgNonincludableNamespaces ) ) {
3165 $found = false; # access denied
3166 wfDebug( __METHOD__
.": template inclusion denied for " . $title->getPrefixedDBkey() );
3168 list( $text, $title ) = $this->getTemplateDom( $title );
3169 if ( $text !== false ) {
3175 # If the title is valid but undisplayable, make a link to it
3176 if ( !$found && ( $this->ot
['html'] ||
$this->ot
['pre'] ) ) {
3177 $text = "[[:$titleText]]";
3180 } elseif ( $title->isTrans() ) {
3181 # Interwiki transclusion
3182 if ( $this->ot
['html'] && !$forceRawInterwiki ) {
3183 $text = $this->interwikiTransclude( $title, 'render' );
3186 $text = $this->interwikiTransclude( $title, 'raw' );
3187 # Preprocess it like a template
3188 $text = $this->preprocessToDom( $text, self
::PTD_FOR_INCLUSION
);
3194 # Do infinite loop check
3195 # This has to be done after redirect resolution to avoid infinite loops via redirects
3196 if ( !$frame->loopCheck( $title ) ) {
3198 $text = '<span class="error">' . wfMsgForContent( 'parser-template-loop-warning', $titleText ) . '</span>';
3199 wfDebug( __METHOD__
.": template loop broken at '$titleText'\n" );
3201 wfProfileOut( __METHOD__
. '-loadtpl' );
3204 # If we haven't found text to substitute by now, we're done
3205 # Recover the source wikitext and return it
3207 $text = $frame->virtualBracketedImplode( '{{', '|', '}}', $titleWithSpaces, $args );
3208 wfProfileOut( __METHOD__
);
3209 return array( 'object' => $text );
3212 # Expand DOM-style return values in a child frame
3213 if ( $isChildObj ) {
3214 # Clean up argument array
3215 $newFrame = $frame->newChild( $args, $title );
3218 $text = $newFrame->expand( $text, PPFrame
::RECOVER_ORIG
);
3219 } elseif ( $titleText !== false && $newFrame->isEmpty() ) {
3220 # Expansion is eligible for the empty-frame cache
3221 if ( isset( $this->mTplExpandCache
[$titleText] ) ) {
3222 $text = $this->mTplExpandCache
[$titleText];
3224 $text = $newFrame->expand( $text );
3225 $this->mTplExpandCache
[$titleText] = $text;
3228 # Uncached expansion
3229 $text = $newFrame->expand( $text );
3232 if ( $isLocalObj && $nowiki ) {
3233 $text = $frame->expand( $text, PPFrame
::RECOVER_ORIG
);
3234 $isLocalObj = false;
3237 # Replace raw HTML by a placeholder
3238 # Add a blank line preceding, to prevent it from mucking up
3239 # immediately preceding headings
3241 $text = "\n\n" . $this->insertStripItem( $text );
3242 } elseif ( $nowiki && ( $this->ot
['html'] ||
$this->ot
['pre'] ) ) {
3243 # Escape nowiki-style return values
3244 $text = wfEscapeWikiText( $text );
3245 } elseif ( is_string( $text )
3246 && !$piece['lineStart']
3247 && preg_match( '/^(?:{\\||:|;|#|\*)/', $text ) )
3249 # Bug 529: if the template begins with a table or block-level
3250 # element, it should be treated as beginning a new line.
3251 # This behaviour is somewhat controversial.
3252 $text = "\n" . $text;
3255 if ( is_string( $text ) && !$this->incrementIncludeSize( 'post-expand', strlen( $text ) ) ) {
3256 # Error, oversize inclusion
3257 if ( $titleText !== false ) {
3258 # Make a working, properly escaped link if possible (bug 23588)
3259 $text = "[[:$titleText]]";
3261 # This will probably not be a working link, but at least it may
3262 # provide some hint of where the problem is
3263 preg_replace( '/^:/', '', $originalTitle );
3264 $text = "[[:$originalTitle]]";
3266 $text .= $this->insertStripItem( '<!-- WARNING: template omitted, post-expand include size too large -->' );
3267 $this->limitationWarn( 'post-expand-template-inclusion' );
3270 if ( $isLocalObj ) {
3271 $ret = array( 'object' => $text );
3273 $ret = array( 'text' => $text );
3276 wfProfileOut( __METHOD__
);
3281 * Get the semi-parsed DOM representation of a template with a given title,
3282 * and its redirect destination title. Cached.
3284 function getTemplateDom( $title ) {
3285 $cacheTitle = $title;
3286 $titleText = $title->getPrefixedDBkey();
3288 if ( isset( $this->mTplRedirCache
[$titleText] ) ) {
3289 list( $ns, $dbk ) = $this->mTplRedirCache
[$titleText];
3290 $title = Title
::makeTitle( $ns, $dbk );
3291 $titleText = $title->getPrefixedDBkey();
3293 if ( isset( $this->mTplDomCache
[$titleText] ) ) {
3294 return array( $this->mTplDomCache
[$titleText], $title );
3297 # Cache miss, go to the database
3298 list( $text, $title ) = $this->fetchTemplateAndTitle( $title );
3300 if ( $text === false ) {
3301 $this->mTplDomCache
[$titleText] = false;
3302 return array( false, $title );
3305 $dom = $this->preprocessToDom( $text, self
::PTD_FOR_INCLUSION
);
3306 $this->mTplDomCache
[ $titleText ] = $dom;
3308 if ( !$title->equals( $cacheTitle ) ) {
3309 $this->mTplRedirCache
[$cacheTitle->getPrefixedDBkey()] =
3310 array( $title->getNamespace(),$cdb = $title->getDBkey() );
3313 return array( $dom, $title );
3317 * Fetch the unparsed text of a template and register a reference to it.
3319 function fetchTemplateAndTitle( $title ) {
3320 $templateCb = $this->mOptions
->getTemplateCallback(); # Defaults to Parser::statelessFetchTemplate()
3321 $stuff = call_user_func( $templateCb, $title, $this );
3322 $text = $stuff['text'];
3323 $finalTitle = isset( $stuff['finalTitle'] ) ?
$stuff['finalTitle'] : $title;
3324 if ( isset( $stuff['deps'] ) ) {
3325 foreach ( $stuff['deps'] as $dep ) {
3326 $this->mOutput
->addTemplate( $dep['title'], $dep['page_id'], $dep['rev_id'] );
3329 return array( $text, $finalTitle );
3332 function fetchTemplate( $title ) {
3333 $rv = $this->fetchTemplateAndTitle( $title );
3338 * Static function to get a template
3339 * Can be overridden via ParserOptions::setTemplateCallback().
3341 static function statelessFetchTemplate( $title, $parser=false ) {
3342 $text = $skip = false;
3343 $finalTitle = $title;
3346 # Loop to fetch the article, with up to 1 redirect
3347 for ( $i = 0; $i < 2 && is_object( $title ); $i++
) {
3348 # Give extensions a chance to select the revision instead
3349 $id = false; # Assume current
3350 wfRunHooks( 'BeforeParserFetchTemplateAndtitle', array( $parser, &$title, &$skip, &$id ) );
3356 'page_id' => $title->getArticleID(),
3360 $rev = $id ? Revision
::newFromId( $id ) : Revision
::newFromTitle( $title );
3361 $rev_id = $rev ?
$rev->getId() : 0;
3362 # If there is no current revision, there is no page
3363 if ( $id === false && !$rev ) {
3364 $linkCache = LinkCache
::singleton();
3365 $linkCache->addBadLinkObj( $title );
3370 'page_id' => $title->getArticleID(),
3371 'rev_id' => $rev_id );
3374 $text = $rev->getText();
3375 } elseif ( $title->getNamespace() == NS_MEDIAWIKI
) {
3377 $message = $wgContLang->lcfirst( $title->getText() );
3378 $text = wfMsgForContentNoTrans( $message );
3379 if ( wfEmptyMsg( $message, $text ) ) {
3386 if ( $text === false ) {
3390 $finalTitle = $title;
3391 $title = Title
::newFromRedirect( $text );
3395 'finalTitle' => $finalTitle,
3400 * Transclude an interwiki link.
3402 function interwikiTransclude( $title, $action ) {
3403 global $wgEnableScaryTranscluding;
3405 if ( !$wgEnableScaryTranscluding ) {
3406 return wfMsgForContent('scarytranscludedisabled');
3409 $url = $title->getFullUrl( "action=$action" );
3411 if ( strlen( $url ) > 255 ) {
3412 return wfMsgForContent( 'scarytranscludetoolong' );
3414 return $this->fetchScaryTemplateMaybeFromCache( $url );
3417 function fetchScaryTemplateMaybeFromCache( $url ) {
3418 global $wgTranscludeCacheExpiry;
3419 $dbr = wfGetDB( DB_SLAVE
);
3420 $tsCond = $dbr->timestamp( time() - $wgTranscludeCacheExpiry );
3421 $obj = $dbr->selectRow( 'transcache', array('tc_time', 'tc_contents' ),
3422 array( 'tc_url' => $url, "tc_time >= " . $dbr->addQuotes( $tsCond ) ) );
3424 return $obj->tc_contents
;
3427 $text = Http
::get( $url );
3429 return wfMsgForContent( 'scarytranscludefailed', $url );
3432 $dbw = wfGetDB( DB_MASTER
);
3433 $dbw->replace( 'transcache', array('tc_url'), array(
3435 'tc_time' => $dbw->timestamp( time() ),
3436 'tc_contents' => $text)
3443 * Triple brace replacement -- used for template arguments
3446 function argSubstitution( $piece, $frame ) {
3447 wfProfileIn( __METHOD__
);
3450 $parts = $piece['parts'];
3451 $nameWithSpaces = $frame->expand( $piece['title'] );
3452 $argName = trim( $nameWithSpaces );
3454 $text = $frame->getArgument( $argName );
3455 if ( $text === false && $parts->getLength() > 0
3459 ||
( $this->ot
['wiki'] && $frame->isTemplate() )
3462 # No match in frame, use the supplied default
3463 $object = $parts->item( 0 )->getChildren();
3465 if ( !$this->incrementIncludeSize( 'arg', strlen( $text ) ) ) {
3466 $error = '<!-- WARNING: argument omitted, expansion size too large -->';
3467 $this->limitationWarn( 'post-expand-template-argument' );
3470 if ( $text === false && $object === false ) {
3472 $object = $frame->virtualBracketedImplode( '{{{', '|', '}}}', $nameWithSpaces, $parts );
3474 if ( $error !== false ) {
3477 if ( $object !== false ) {
3478 $ret = array( 'object' => $object );
3480 $ret = array( 'text' => $text );
3483 wfProfileOut( __METHOD__
);
3488 * Return the text to be used for a given extension tag.
3489 * This is the ghost of strip().
3491 * @param $params Associative array of parameters:
3492 * name PPNode for the tag name
3493 * attr PPNode for unparsed text where tag attributes are thought to be
3494 * attributes Optional associative array of parsed attributes
3495 * inner Contents of extension element
3496 * noClose Original text did not have a close tag
3497 * @param $frame PPFrame
3499 function extensionSubstitution( $params, $frame ) {
3500 $name = $frame->expand( $params['name'] );
3501 $attrText = !isset( $params['attr'] ) ?
null : $frame->expand( $params['attr'] );
3502 $content = !isset( $params['inner'] ) ?
null : $frame->expand( $params['inner'] );
3503 $marker = "{$this->mUniqPrefix}-$name-" . sprintf( '%08X', $this->mMarkerIndex++
) . self
::MARKER_SUFFIX
;
3505 $isFunctionTag = isset( $this->mFunctionTagHooks
[strtolower($name)] ) &&
3506 ( $this->ot
['html'] ||
$this->ot
['pre'] );
3507 if ( $isFunctionTag ) {
3508 $markerType = 'none';
3510 $markerType = 'general';
3512 if ( $this->ot
['html'] ||
$isFunctionTag ) {
3513 $name = strtolower( $name );
3514 $attributes = Sanitizer
::decodeTagAttributes( $attrText );
3515 if ( isset( $params['attributes'] ) ) {
3516 $attributes = $attributes +
$params['attributes'];
3519 if ( isset( $this->mTagHooks
[$name] ) ) {
3520 # Workaround for PHP bug 35229 and similar
3521 if ( !is_callable( $this->mTagHooks
[$name] ) ) {
3522 throw new MWException( "Tag hook for $name is not callable\n" );
3524 $output = call_user_func_array( $this->mTagHooks
[$name],
3525 array( $content, $attributes, $this, $frame ) );
3526 } elseif ( isset( $this->mFunctionTagHooks
[$name] ) ) {
3527 list( $callback, $flags ) = $this->mFunctionTagHooks
[$name];
3528 if ( !is_callable( $callback ) ) {
3529 throw new MWException( "Tag hook for $name is not callable\n" );
3532 $output = call_user_func_array( $callback, array( &$this, $frame, $content, $attributes ) );
3534 $output = '<span class="error">Invalid tag extension name: ' .
3535 htmlspecialchars( $name ) . '</span>';
3538 if ( is_array( $output ) ) {
3539 # Extract flags to local scope (to override $markerType)
3541 $output = $flags[0];
3546 if ( is_null( $attrText ) ) {
3549 if ( isset( $params['attributes'] ) ) {
3550 foreach ( $params['attributes'] as $attrName => $attrValue ) {
3551 $attrText .= ' ' . htmlspecialchars( $attrName ) . '="' .
3552 htmlspecialchars( $attrValue ) . '"';
3555 if ( $content === null ) {
3556 $output = "<$name$attrText/>";
3558 $close = is_null( $params['close'] ) ?
'' : $frame->expand( $params['close'] );
3559 $output = "<$name$attrText>$content$close";
3563 if ( $markerType === 'none' ) {
3565 } elseif ( $markerType === 'nowiki' ) {
3566 $this->mStripState
->nowiki
->setPair( $marker, $output );
3567 } elseif ( $markerType === 'general' ) {
3568 $this->mStripState
->general
->setPair( $marker, $output );
3570 throw new MWException( __METHOD__
.': invalid marker type' );
3576 * Increment an include size counter
3578 * @param $type String: the type of expansion
3579 * @param $size Integer: the size of the text
3580 * @return Boolean: false if this inclusion would take it over the maximum, true otherwise
3582 function incrementIncludeSize( $type, $size ) {
3583 if ( $this->mIncludeSizes
[$type] +
$size > $this->mOptions
->getMaxIncludeSize( $type ) ) {
3586 $this->mIncludeSizes
[$type] +
= $size;
3592 * Increment the expensive function count
3594 * @return Boolean: false if the limit has been exceeded
3596 function incrementExpensiveFunctionCount() {
3597 global $wgExpensiveParserFunctionLimit;
3598 $this->mExpensiveFunctionCount++
;
3599 if ( $this->mExpensiveFunctionCount
<= $wgExpensiveParserFunctionLimit ) {
3606 * Strip double-underscore items like __NOGALLERY__ and __NOTOC__
3607 * Fills $this->mDoubleUnderscores, returns the modified text
3609 function doDoubleUnderscore( $text ) {
3610 wfProfileIn( __METHOD__
);
3612 # The position of __TOC__ needs to be recorded
3613 $mw = MagicWord
::get( 'toc' );
3614 if ( $mw->match( $text ) ) {
3615 $this->mShowToc
= true;
3616 $this->mForceTocPosition
= true;
3618 # Set a placeholder. At the end we'll fill it in with the TOC.
3619 $text = $mw->replace( '<!--MWTOC-->', $text, 1 );
3621 # Only keep the first one.
3622 $text = $mw->replace( '', $text );
3625 # Now match and remove the rest of them
3626 $mwa = MagicWord
::getDoubleUnderscoreArray();
3627 $this->mDoubleUnderscores
= $mwa->matchAndRemove( $text );
3629 if ( isset( $this->mDoubleUnderscores
['nogallery'] ) ) {
3630 $this->mOutput
->mNoGallery
= true;
3632 if ( isset( $this->mDoubleUnderscores
['notoc'] ) && !$this->mForceTocPosition
) {
3633 $this->mShowToc
= false;
3635 if ( isset( $this->mDoubleUnderscores
['hiddencat'] ) && $this->mTitle
->getNamespace() == NS_CATEGORY
) {
3636 $this->addTrackingCategory( 'hidden-category-category' );
3638 # (bug 8068) Allow control over whether robots index a page.
3640 # FIXME (bug 14899): __INDEX__ always overrides __NOINDEX__ here! This
3641 # is not desirable, the last one on the page should win.
3642 if ( isset( $this->mDoubleUnderscores
['noindex'] ) && $this->mTitle
->canUseNoindex() ) {
3643 $this->mOutput
->setIndexPolicy( 'noindex' );
3644 $this->addTrackingCategory( 'noindex-category' );
3646 if ( isset( $this->mDoubleUnderscores
['index'] ) && $this->mTitle
->canUseNoindex() ) {
3647 $this->mOutput
->setIndexPolicy( 'index' );
3648 $this->addTrackingCategory( 'index-category' );
3651 # Cache all double underscores in the database
3652 foreach ( $this->mDoubleUnderscores
as $key => $val ) {
3653 $this->mOutput
->setProperty( $key, '' );
3656 wfProfileOut( __METHOD__
);
3661 * Add a tracking category, getting the title from a system message,
3662 * or print a debug message if the title is invalid.
3664 * @param $msg String: message key
3665 * @return Boolean: whether the addition was successful
3667 protected function addTrackingCategory( $msg ) {
3668 $cat = wfMsgForContent( $msg );
3670 # Allow tracking categories to be disabled by setting them to "-"
3671 if ( $cat === '-' ) {
3675 $containerCategory = Title
::makeTitleSafe( NS_CATEGORY
, $cat );
3676 if ( $containerCategory ) {
3677 $this->mOutput
->addCategory( $containerCategory->getDBkey(), $this->getDefaultSort() );
3680 wfDebug( __METHOD__
.": [[MediaWiki:$msg]] is not a valid title!\n" );
3686 * This function accomplishes several tasks:
3687 * 1) Auto-number headings if that option is enabled
3688 * 2) Add an [edit] link to sections for users who have enabled the option and can edit the page
3689 * 3) Add a Table of contents on the top for users who have enabled the option
3690 * 4) Auto-anchor headings
3692 * It loops through all headlines, collects the necessary data, then splits up the
3693 * string and re-inserts the newly formatted headlines.
3695 * @param $text String
3696 * @param $origText String: original, untouched wikitext
3697 * @param $isMain Boolean
3700 function formatHeadings( $text, $origText, $isMain=true ) {
3701 global $wgMaxTocLevel, $wgContLang, $wgHtml5, $wgExperimentalHtmlIds;
3703 $doNumberHeadings = $this->mOptions
->getNumberHeadings();
3705 # Inhibit editsection links if requested in the page
3706 if ( isset( $this->mDoubleUnderscores
['noeditsection'] ) ) {
3709 $showEditLink = $this->mOptions
->getEditSection();
3712 # Get all headlines for numbering them and adding funky stuff like [edit]
3713 # links - this is for later, but we need the number of headlines right now
3715 $numMatches = preg_match_all( '/<H(?P<level>[1-6])(?P<attrib>.*?'.'>)(?P<header>.*?)<\/H[1-6] *>/i', $text, $matches );
3717 # if there are fewer than 4 headlines in the article, do not show TOC
3718 # unless it's been explicitly enabled.
3719 $enoughToc = $this->mShowToc
&&
3720 ( ( $numMatches >= 4 ) ||
$this->mForceTocPosition
);
3722 # Allow user to stipulate that a page should have a "new section"
3723 # link added via __NEWSECTIONLINK__
3724 if ( isset( $this->mDoubleUnderscores
['newsectionlink'] ) ) {
3725 $this->mOutput
->setNewSection( true );
3728 # Allow user to remove the "new section"
3729 # link via __NONEWSECTIONLINK__
3730 if ( isset( $this->mDoubleUnderscores
['nonewsectionlink'] ) ) {
3731 $this->mOutput
->hideNewSection( true );
3734 # if the string __FORCETOC__ (not case-sensitive) occurs in the HTML,
3735 # override above conditions and always show TOC above first header
3736 if ( isset( $this->mDoubleUnderscores
['forcetoc'] ) ) {
3737 $this->mShowToc
= true;
3741 # We need this to perform operations on the HTML
3742 $sk = $this->mOptions
->getSkin( $this->mTitle
);
3748 # Ugh .. the TOC should have neat indentation levels which can be
3749 # passed to the skin functions. These are determined here
3753 $sublevelCount = array();
3754 $levelCount = array();
3760 $markerRegex = "{$this->mUniqPrefix}-h-(\d+)-" . self
::MARKER_SUFFIX
;
3761 $baseTitleText = $this->mTitle
->getPrefixedDBkey();
3762 $oldType = $this->mOutputType
;
3763 $this->setOutputType( self
::OT_WIKI
);
3764 $frame = $this->getPreprocessor()->newFrame();
3765 $root = $this->preprocessToDom( $origText );
3766 $node = $root->getFirstChild();
3770 foreach ( $matches[3] as $headline ) {
3771 $isTemplate = false;
3773 $sectionIndex = false;
3775 $markerMatches = array();
3776 if ( preg_match("/^$markerRegex/", $headline, $markerMatches ) ) {
3777 $serial = $markerMatches[1];
3778 list( $titleText, $sectionIndex ) = $this->mHeadings
[$serial];
3779 $isTemplate = ( $titleText != $baseTitleText );
3780 $headline = preg_replace( "/^$markerRegex/", "", $headline );
3784 $prevlevel = $level;
3786 $level = $matches[1][$headlineCount];
3788 if ( $level > $prevlevel ) {
3789 # Increase TOC level
3791 $sublevelCount[$toclevel] = 0;
3792 if ( $toclevel<$wgMaxTocLevel ) {
3793 $prevtoclevel = $toclevel;
3794 $toc .= $sk->tocIndent();
3797 } elseif ( $level < $prevlevel && $toclevel > 1 ) {
3798 # Decrease TOC level, find level to jump to
3800 for ( $i = $toclevel; $i > 0; $i-- ) {
3801 if ( $levelCount[$i] == $level ) {
3802 # Found last matching level
3805 } elseif ( $levelCount[$i] < $level ) {
3806 # Found first matching level below current level
3814 if ( $toclevel<$wgMaxTocLevel ) {
3815 if ( $prevtoclevel < $wgMaxTocLevel ) {
3816 # Unindent only if the previous toc level was shown :p
3817 $toc .= $sk->tocUnindent( $prevtoclevel - $toclevel );
3818 $prevtoclevel = $toclevel;
3820 $toc .= $sk->tocLineEnd();
3824 # No change in level, end TOC line
3825 if ( $toclevel<$wgMaxTocLevel ) {
3826 $toc .= $sk->tocLineEnd();
3830 $levelCount[$toclevel] = $level;
3832 # count number of headlines for each level
3833 @$sublevelCount[$toclevel]++
;
3835 for( $i = 1; $i <= $toclevel; $i++
) {
3836 if ( !empty( $sublevelCount[$i] ) ) {
3840 $numbering .= $wgContLang->formatNum( $sublevelCount[$i] );
3845 # The safe header is a version of the header text safe to use for links
3846 # Avoid insertion of weird stuff like <math> by expanding the relevant sections
3847 $safeHeadline = $this->mStripState
->unstripBoth( $headline );
3849 # Remove link placeholders by the link text.
3850 # <!--LINK number-->
3852 # link text with suffix
3853 $safeHeadline = $this->replaceLinkHoldersText( $safeHeadline );
3855 # Strip out HTML (other than plain <sup> and <sub>: bug 8393)
3856 $tocline = preg_replace(
3857 array( '#<(?!/?(sup|sub)).*?'.'>#', '#<(/?(sup|sub)).*?'.'>#' ),
3858 array( '', '<$1>' ),
3861 $tocline = trim( $tocline );
3863 # For the anchor, strip out HTML-y stuff period
3864 $safeHeadline = preg_replace( '/<.*?'.'>/', '', $safeHeadline );
3865 $safeHeadline = Sanitizer
::normalizeSectionNameWhitespace( $safeHeadline );
3867 # Save headline for section edit hint before it's escaped
3868 $headlineHint = $safeHeadline;
3870 if ( $wgHtml5 && $wgExperimentalHtmlIds ) {
3871 # For reverse compatibility, provide an id that's
3872 # HTML4-compatible, like we used to.
3874 # It may be worth noting, academically, that it's possible for
3875 # the legacy anchor to conflict with a non-legacy headline
3876 # anchor on the page. In this case likely the "correct" thing
3877 # would be to either drop the legacy anchors or make sure
3878 # they're numbered first. However, this would require people
3879 # to type in section names like "abc_.D7.93.D7.90.D7.A4"
3880 # manually, so let's not bother worrying about it.
3881 $legacyHeadline = Sanitizer
::escapeId( $safeHeadline,
3882 array( 'noninitial', 'legacy' ) );
3883 $safeHeadline = Sanitizer
::escapeId( $safeHeadline );
3885 if ( $legacyHeadline == $safeHeadline ) {
3886 # No reason to have both (in fact, we can't)
3887 $legacyHeadline = false;
3890 $legacyHeadline = false;
3891 $safeHeadline = Sanitizer
::escapeId( $safeHeadline,
3895 # HTML names must be case-insensitively unique (bug 10721). FIXME:
3896 # Does this apply to Unicode characters? Because we aren't
3897 # handling those here.
3898 $arrayKey = strtolower( $safeHeadline );
3899 if ( $legacyHeadline === false ) {
3900 $legacyArrayKey = false;
3902 $legacyArrayKey = strtolower( $legacyHeadline );
3905 # count how many in assoc. array so we can track dupes in anchors
3906 if ( isset( $refers[$arrayKey] ) ) {
3907 $refers[$arrayKey]++
;
3909 $refers[$arrayKey] = 1;
3911 if ( isset( $refers[$legacyArrayKey] ) ) {
3912 $refers[$legacyArrayKey]++
;
3914 $refers[$legacyArrayKey] = 1;
3917 # Don't number the heading if it is the only one (looks silly)
3918 if ( $doNumberHeadings && count( $matches[3] ) > 1) {
3919 # the two are different if the line contains a link
3920 $headline = $numbering . ' ' . $headline;
3923 # Create the anchor for linking from the TOC to the section
3924 $anchor = $safeHeadline;
3925 $legacyAnchor = $legacyHeadline;
3926 if ( $refers[$arrayKey] > 1 ) {
3927 $anchor .= '_' . $refers[$arrayKey];
3929 if ( $legacyHeadline !== false && $refers[$legacyArrayKey] > 1 ) {
3930 $legacyAnchor .= '_' . $refers[$legacyArrayKey];
3932 if ( $enoughToc && ( !isset( $wgMaxTocLevel ) ||
$toclevel < $wgMaxTocLevel ) ) {
3933 $toc .= $sk->tocLine( $anchor, $tocline,
3934 $numbering, $toclevel, ( $isTemplate ?
false : $sectionIndex ) );
3937 # Add the section to the section tree
3938 # Find the DOM node for this header
3939 while ( $node && !$isTemplate ) {
3940 if ( $node->getName() === 'h' ) {
3941 $bits = $node->splitHeading();
3942 if ( $bits['i'] == $sectionIndex )
3945 $byteOffset +
= mb_strlen( $this->mStripState
->unstripBoth(
3946 $frame->expand( $node, PPFrame
::RECOVER_ORIG
) ) );
3947 $node = $node->getNextSibling();
3950 'toclevel' => $toclevel,
3953 'number' => $numbering,
3954 'index' => ( $isTemplate ?
'T-' : '' ) . $sectionIndex,
3955 'fromtitle' => $titleText,
3956 'byteoffset' => ( $isTemplate ?
null : $byteOffset ),
3957 'anchor' => $anchor,
3960 # give headline the correct <h#> tag
3961 if ( $showEditLink && $sectionIndex !== false ) {
3962 if ( $isTemplate ) {
3963 # Put a T flag in the section identifier, to indicate to extractSections()
3964 # that sections inside <includeonly> should be counted.
3965 $editlink = $sk->doEditSectionLink( Title
::newFromText( $titleText ), "T-$sectionIndex", null, $this->mOptions
->getUserLang() );
3967 $editlink = $sk->doEditSectionLink( $this->mTitle
, $sectionIndex, $headlineHint, $this->mOptions
->getUserLang() );
3972 $head[$headlineCount] = $sk->makeHeadline( $level,
3973 $matches['attrib'][$headlineCount], $anchor, $headline,
3974 $editlink, $legacyAnchor );
3979 $this->setOutputType( $oldType );
3981 # Never ever show TOC if no headers
3982 if ( $numVisible < 1 ) {
3987 if ( $prevtoclevel > 0 && $prevtoclevel < $wgMaxTocLevel ) {
3988 $toc .= $sk->tocUnindent( $prevtoclevel - 1 );
3990 $toc = $sk->tocList( $toc );
3991 $this->mOutput
->setTOCHTML( $toc );
3995 $this->mOutput
->setSections( $tocraw );
3998 # split up and insert constructed headlines
4000 $blocks = preg_split( '/<H[1-6].*?' . '>.*?<\/H[1-6]>/i', $text );
4003 foreach ( $blocks as $block ) {
4004 if ( $showEditLink && $headlineCount > 0 && $i == 0 && $block !== "\n" ) {
4005 # This is the [edit] link that appears for the top block of text when
4006 # section editing is enabled
4008 # Disabled because it broke block formatting
4009 # For example, a bullet point in the top line
4010 # $full .= $sk->editSectionLink(0);
4013 if ( $enoughToc && !$i && $isMain && !$this->mForceTocPosition
) {
4014 # Top anchor now in skin
4018 if ( !empty( $head[$i] ) ) {
4023 if ( $this->mForceTocPosition
) {
4024 return str_replace( '<!--MWTOC-->', $toc, $full );
4031 * Transform wiki markup when saving a page by doing \r\n -> \n
4032 * conversion, substitting signatures, {{subst:}} templates, etc.
4034 * @param $text String: the text to transform
4035 * @param &$title Title: the Title object for the current article
4036 * @param $user User: the User object describing the current user
4037 * @param $options ParserOptions: parsing options
4038 * @param $clearState Boolean: whether to clear the parser state first
4039 * @return String: the altered wiki markup
4041 public function preSaveTransform( $text, Title
$title, $user, $options, $clearState = true ) {
4042 $options->resetUsage();
4043 $this->mOptions
= $options;
4044 $this->setTitle( $title );
4045 $this->setOutputType( self
::OT_WIKI
);
4047 if ( $clearState ) {
4048 $this->clearState();
4054 $text = str_replace( array_keys( $pairs ), array_values( $pairs ), $text );
4055 $text = $this->pstPass2( $text, $user );
4056 $text = $this->mStripState
->unstripBoth( $text );
4061 * Pre-save transform helper function
4064 function pstPass2( $text, $user ) {
4065 global $wgContLang, $wgLocaltimezone;
4067 # Note: This is the timestamp saved as hardcoded wikitext to
4068 # the database, we use $wgContLang here in order to give
4069 # everyone the same signature and use the default one rather
4070 # than the one selected in each user's preferences.
4071 # (see also bug 12815)
4072 $ts = $this->mOptions
->getTimestamp();
4073 if ( isset( $wgLocaltimezone ) ) {
4074 $tz = $wgLocaltimezone;
4076 $tz = date_default_timezone_get();
4079 $unixts = wfTimestamp( TS_UNIX
, $ts );
4080 $oldtz = date_default_timezone_get();
4081 date_default_timezone_set( $tz );
4082 $ts = date( 'YmdHis', $unixts );
4083 $tzMsg = date( 'T', $unixts ); # might vary on DST changeover!
4085 # Allow translation of timezones through wiki. date() can return
4086 # whatever crap the system uses, localised or not, so we cannot
4087 # ship premade translations.
4088 $key = 'timezone-' . strtolower( trim( $tzMsg ) );
4089 $value = wfMsgForContent( $key );
4090 if ( !wfEmptyMsg( $key, $value ) ) {
4094 date_default_timezone_set( $oldtz );
4096 $d = $wgContLang->timeanddate( $ts, false, false ) . " ($tzMsg)";
4098 # Variable replacement
4099 # Because mOutputType is OT_WIKI, this will only process {{subst:xxx}} type tags
4100 $text = $this->replaceVariables( $text );
4103 $sigText = $this->getUserSig( $user );
4104 $text = strtr( $text, array(
4106 '~~~~' => "$sigText $d",
4110 # Context links: [[|name]] and [[name (context)|]]
4111 global $wgLegalTitleChars;
4112 $tc = "[$wgLegalTitleChars]";
4113 $nc = '[ _0-9A-Za-z\x80-\xff-]'; # Namespaces can use non-ascii!
4115 $p1 = "/\[\[(:?$nc+:|:|)($tc+?)( \\($tc+\\))\\|]]/"; # [[ns:page (context)|]]
4116 $p4 = "/\[\[(:?$nc+:|:|)($tc+?)(($tc+))\\|]]/"; # [[ns:page(context)|]]
4117 $p3 = "/\[\[(:?$nc+:|:|)($tc+?)( \\($tc+\\)|)(, $tc+|)\\|]]/"; # [[ns:page (context), context|]]
4118 $p2 = "/\[\[\\|($tc+)]]/"; # [[|page]]
4120 # try $p1 first, to turn "[[A, B (C)|]]" into "[[A, B (C)|A, B]]"
4121 $text = preg_replace( $p1, '[[\\1\\2\\3|\\2]]', $text );
4122 $text = preg_replace( $p4, '[[\\1\\2\\3|\\2]]', $text );
4123 $text = preg_replace( $p3, '[[\\1\\2\\3\\4|\\2]]', $text );
4125 $t = $this->mTitle
->getText();
4127 if ( preg_match( "/^($nc+:|)$tc+?( \\($tc+\\))$/", $t, $m ) ) {
4128 $text = preg_replace( $p2, "[[$m[1]\\1$m[2]|\\1]]", $text );
4129 } elseif ( preg_match( "/^($nc+:|)$tc+?(, $tc+|)$/", $t, $m ) && "$m[1]$m[2]" != '' ) {
4130 $text = preg_replace( $p2, "[[$m[1]\\1$m[2]|\\1]]", $text );
4132 # if there's no context, don't bother duplicating the title
4133 $text = preg_replace( $p2, '[[\\1]]', $text );
4136 # Trim trailing whitespace
4137 $text = rtrim( $text );
4143 * Fetch the user's signature text, if any, and normalize to
4144 * validated, ready-to-insert wikitext.
4145 * If you have pre-fetched the nickname or the fancySig option, you can
4146 * specify them here to save a database query.
4149 * @param $nickname String: nickname to use or false to use user's default nickname
4150 * @param $fancySig Boolean: whether the nicknname is the complete signature
4151 * or null to use default value
4154 function getUserSig( &$user, $nickname = false, $fancySig = null ) {
4155 global $wgMaxSigChars;
4157 $username = $user->getName();
4159 # If not given, retrieve from the user object.
4160 if ( $nickname === false )
4161 $nickname = $user->getOption( 'nickname' );
4163 if ( is_null( $fancySig ) ) {
4164 $fancySig = $user->getBoolOption( 'fancysig' );
4167 $nickname = $nickname == null ?
$username : $nickname;
4169 if ( mb_strlen( $nickname ) > $wgMaxSigChars ) {
4170 $nickname = $username;
4171 wfDebug( __METHOD__
. ": $username has overlong signature.\n" );
4172 } elseif ( $fancySig !== false ) {
4173 # Sig. might contain markup; validate this
4174 if ( $this->validateSig( $nickname ) !== false ) {
4175 # Validated; clean up (if needed) and return it
4176 return $this->cleanSig( $nickname, true );
4178 # Failed to validate; fall back to the default
4179 $nickname = $username;
4180 wfDebug( __METHOD__
.": $username has bad XML tags in signature.\n" );
4184 # Make sure nickname doesnt get a sig in a sig
4185 $nickname = $this->cleanSigInSig( $nickname );
4187 # If we're still here, make it a link to the user page
4188 $userText = wfEscapeWikiText( $username );
4189 $nickText = wfEscapeWikiText( $nickname );
4190 if ( $user->isAnon() ) {
4191 return wfMsgExt( 'signature-anon', array( 'content', 'parsemag' ), $userText, $nickText );
4193 return wfMsgExt( 'signature', array( 'content', 'parsemag' ), $userText, $nickText );
4198 * Check that the user's signature contains no bad XML
4200 * @param $text String
4201 * @return mixed An expanded string, or false if invalid.
4203 function validateSig( $text ) {
4204 return( Xml
::isWellFormedXmlFragment( $text ) ?
$text : false );
4208 * Clean up signature text
4210 * 1) Strip ~~~, ~~~~ and ~~~~~ out of signatures @see cleanSigInSig
4211 * 2) Substitute all transclusions
4213 * @param $text String
4214 * @param $parsing Whether we're cleaning (preferences save) or parsing
4215 * @return String: signature text
4217 function cleanSig( $text, $parsing = false ) {
4220 $this->clearState();
4221 $this->setTitle( $wgTitle );
4222 $this->mOptions
= new ParserOptions
;
4223 $this->setOutputType
= self
::OT_PREPROCESS
;
4226 # Option to disable this feature
4227 if ( !$this->mOptions
->getCleanSignatures() ) {
4231 # FIXME: regex doesn't respect extension tags or nowiki
4232 # => Move this logic to braceSubstitution()
4233 $substWord = MagicWord
::get( 'subst' );
4234 $substRegex = '/\{\{(?!(?:' . $substWord->getBaseRegex() . '))/x' . $substWord->getRegexCase();
4235 $substText = '{{' . $substWord->getSynonym( 0 );
4237 $text = preg_replace( $substRegex, $substText, $text );
4238 $text = $this->cleanSigInSig( $text );
4239 $dom = $this->preprocessToDom( $text );
4240 $frame = $this->getPreprocessor()->newFrame();
4241 $text = $frame->expand( $dom );
4244 $text = $this->mStripState
->unstripBoth( $text );
4251 * Strip ~~~, ~~~~ and ~~~~~ out of signatures
4253 * @param $text String
4254 * @return String: signature text with /~{3,5}/ removed
4256 function cleanSigInSig( $text ) {
4257 $text = preg_replace( '/~{3,5}/', '', $text );
4262 * Set up some variables which are usually set up in parse()
4263 * so that an external function can call some class members with confidence
4265 public function startExternalParse( &$title, $options, $outputType, $clearState = true ) {
4266 $this->setTitle( $title );
4267 $options->resetUsage();
4268 $this->mOptions
= $options;
4269 $this->setOutputType( $outputType );
4270 if ( $clearState ) {
4271 $this->clearState();
4276 * Wrapper for preprocess()
4278 * @param $text String: the text to preprocess
4279 * @param $options ParserOptions: options
4282 public function transformMsg( $text, $options ) {
4284 static $executing = false;
4286 # Guard against infinite recursion
4292 wfProfileIn( __METHOD__
);
4293 $text = $this->preprocess( $text, $wgTitle, $options );
4296 wfProfileOut( __METHOD__
);
4301 * Create an HTML-style tag, e.g. <yourtag>special text</yourtag>
4302 * The callback should have the following form:
4303 * function myParserHook( $text, $params, $parser ) { ... }
4305 * Transform and return $text. Use $parser for any required context, e.g. use
4306 * $parser->getTitle() and $parser->getOptions() not $wgTitle or $wgOut->mParserOptions
4308 * @param $tag Mixed: the tag to use, e.g. 'hook' for <hook>
4309 * @param $callback Mixed: the callback function (and object) to use for the tag
4310 * @return The old value of the mTagHooks array associated with the hook
4312 public function setHook( $tag, $callback ) {
4313 $tag = strtolower( $tag );
4314 $oldVal = isset( $this->mTagHooks
[$tag] ) ?
$this->mTagHooks
[$tag] : null;
4315 $this->mTagHooks
[$tag] = $callback;
4316 if ( !in_array( $tag, $this->mStripList
) ) {
4317 $this->mStripList
[] = $tag;
4323 function setTransparentTagHook( $tag, $callback ) {
4324 $tag = strtolower( $tag );
4325 $oldVal = isset( $this->mTransparentTagHooks
[$tag] ) ?
$this->mTransparentTagHooks
[$tag] : null;
4326 $this->mTransparentTagHooks
[$tag] = $callback;
4332 * Remove all tag hooks
4334 function clearTagHooks() {
4335 $this->mTagHooks
= array();
4336 $this->mStripList
= $this->mDefaultStripList
;
4340 * Create a function, e.g. {{sum:1|2|3}}
4341 * The callback function should have the form:
4342 * function myParserFunction( &$parser, $arg1, $arg2, $arg3 ) { ... }
4344 * Or with SFH_OBJECT_ARGS:
4345 * function myParserFunction( $parser, $frame, $args ) { ... }
4347 * The callback may either return the text result of the function, or an array with the text
4348 * in element 0, and a number of flags in the other elements. The names of the flags are
4349 * specified in the keys. Valid flags are:
4350 * found The text returned is valid, stop processing the template. This
4352 * nowiki Wiki markup in the return value should be escaped
4353 * isHTML The returned text is HTML, armour it against wikitext transformation
4355 * @param $id String: The magic word ID
4356 * @param $callback Mixed: the callback function (and object) to use
4357 * @param $flags Integer: a combination of the following flags:
4358 * SFH_NO_HASH No leading hash, i.e. {{plural:...}} instead of {{#if:...}}
4360 * SFH_OBJECT_ARGS Pass the template arguments as PPNode objects instead of text. This
4361 * allows for conditional expansion of the parse tree, allowing you to eliminate dead
4362 * branches and thus speed up parsing. It is also possible to analyse the parse tree of
4363 * the arguments, and to control the way they are expanded.
4365 * The $frame parameter is a PPFrame. This can be used to produce expanded text from the
4366 * arguments, for instance:
4367 * $text = isset( $args[0] ) ? $frame->expand( $args[0] ) : '';
4369 * For technical reasons, $args[0] is pre-expanded and will be a string. This may change in
4370 * future versions. Please call $frame->expand() on it anyway so that your code keeps
4371 * working if/when this is changed.
4373 * If you want whitespace to be trimmed from $args, you need to do it yourself, post-
4376 * Please read the documentation in includes/parser/Preprocessor.php for more information
4377 * about the methods available in PPFrame and PPNode.
4379 * @return The old callback function for this name, if any
4381 public function setFunctionHook( $id, $callback, $flags = 0 ) {
4384 $oldVal = isset( $this->mFunctionHooks
[$id] ) ?
$this->mFunctionHooks
[$id][0] : null;
4385 $this->mFunctionHooks
[$id] = array( $callback, $flags );
4387 # Add to function cache
4388 $mw = MagicWord
::get( $id );
4390 throw new MWException( __METHOD__
.'() expecting a magic word identifier.' );
4392 $synonyms = $mw->getSynonyms();
4393 $sensitive = intval( $mw->isCaseSensitive() );
4395 foreach ( $synonyms as $syn ) {
4397 if ( !$sensitive ) {
4398 $syn = $wgContLang->lc( $syn );
4401 if ( !( $flags & SFH_NO_HASH
) ) {
4404 # Remove trailing colon
4405 if ( substr( $syn, -1, 1 ) === ':' ) {
4406 $syn = substr( $syn, 0, -1 );
4408 $this->mFunctionSynonyms
[$sensitive][$syn] = $id;
4414 * Get all registered function hook identifiers
4418 function getFunctionHooks() {
4419 return array_keys( $this->mFunctionHooks
);
4423 * Create a tag function, e.g. <test>some stuff</test>.
4424 * Unlike tag hooks, tag functions are parsed at preprocessor level.
4425 * Unlike parser functions, their content is not preprocessed.
4427 function setFunctionTagHook( $tag, $callback, $flags ) {
4428 $tag = strtolower( $tag );
4429 $old = isset( $this->mFunctionTagHooks
[$tag] ) ?
4430 $this->mFunctionTagHooks
[$tag] : null;
4431 $this->mFunctionTagHooks
[$tag] = array( $callback, $flags );
4433 if ( !in_array( $tag, $this->mStripList
) ) {
4434 $this->mStripList
[] = $tag;
4441 * FIXME: update documentation. makeLinkObj() is deprecated.
4442 * Replace <!--LINK--> link placeholders with actual links, in the buffer
4443 * Placeholders created in Skin::makeLinkObj()
4444 * Returns an array of link CSS classes, indexed by PDBK.
4446 function replaceLinkHolders( &$text, $options = 0 ) {
4447 return $this->mLinkHolders
->replace( $text );
4451 * Replace <!--LINK--> link placeholders with plain text of links
4452 * (not HTML-formatted).
4454 * @param $text String
4457 function replaceLinkHoldersText( $text ) {
4458 return $this->mLinkHolders
->replaceText( $text );
4462 * Renders an image gallery from a text with one line per image.
4463 * text labels may be given by using |-style alternative text. E.g.
4464 * Image:one.jpg|The number "1"
4465 * Image:tree.jpg|A tree
4466 * given as text will return the HTML of a gallery with two images,
4467 * labeled 'The number "1"' and
4470 function renderImageGallery( $text, $params ) {
4471 $ig = new ImageGallery();
4472 $ig->setContextTitle( $this->mTitle
);
4473 $ig->setShowBytes( false );
4474 $ig->setShowFilename( false );
4475 $ig->setParser( $this );
4476 $ig->setHideBadImages();
4477 $ig->setAttributes( Sanitizer
::validateTagAttributes( $params, 'table' ) );
4478 $ig->useSkin( $this->mOptions
->getSkin( $this->mTitle
) );
4479 $ig->mRevisionId
= $this->mRevisionId
;
4481 if ( isset( $params['showfilename'] ) ) {
4482 $ig->setShowFilename( true );
4484 $ig->setShowFilename( false );
4486 if ( isset( $params['caption'] ) ) {
4487 $caption = $params['caption'];
4488 $caption = htmlspecialchars( $caption );
4489 $caption = $this->replaceInternalLinks( $caption );
4490 $ig->setCaptionHtml( $caption );
4492 if ( isset( $params['perrow'] ) ) {
4493 $ig->setPerRow( $params['perrow'] );
4495 if ( isset( $params['widths'] ) ) {
4496 $ig->setWidths( $params['widths'] );
4498 if ( isset( $params['heights'] ) ) {
4499 $ig->setHeights( $params['heights'] );
4502 wfRunHooks( 'BeforeParserrenderImageGallery', array( &$this, &$ig ) );
4504 $lines = StringUtils
::explode( "\n", $text );
4505 foreach ( $lines as $line ) {
4506 # match lines like these:
4507 # Image:someimage.jpg|This is some image
4509 preg_match( "/^([^|]+)(\\|(.*))?$/", $line, $matches );
4511 if ( count( $matches ) == 0 ) {
4515 if ( strpos( $matches[0], '%' ) !== false ) {
4516 $matches[1] = urldecode( $matches[1] );
4518 $tp = Title
::newFromText( $matches[1] );
4520 if ( is_null( $nt ) ) {
4521 # Bogus title. Ignore these so we don't bomb out later.
4524 if ( isset( $matches[3] ) ) {
4525 $label = $matches[3];
4530 $html = $this->recursiveTagParse( trim( $label ) );
4532 $ig->add( $nt, $html );
4534 # Only add real images (bug #5586)
4535 if ( $nt->getNamespace() == NS_FILE
) {
4536 $this->mOutput
->addImage( $nt->getDBkey() );
4539 return $ig->toHTML();
4542 function getImageParams( $handler ) {
4544 $handlerClass = get_class( $handler );
4548 if ( !isset( $this->mImageParams
[$handlerClass] ) ) {
4549 # Initialise static lists
4550 static $internalParamNames = array(
4551 'horizAlign' => array( 'left', 'right', 'center', 'none' ),
4552 'vertAlign' => array( 'baseline', 'sub', 'super', 'top', 'text-top', 'middle',
4553 'bottom', 'text-bottom' ),
4554 'frame' => array( 'thumbnail', 'manualthumb', 'framed', 'frameless',
4555 'upright', 'border', 'link', 'alt' ),
4557 static $internalParamMap;
4558 if ( !$internalParamMap ) {
4559 $internalParamMap = array();
4560 foreach ( $internalParamNames as $type => $names ) {
4561 foreach ( $names as $name ) {
4562 $magicName = str_replace( '-', '_', "img_$name" );
4563 $internalParamMap[$magicName] = array( $type, $name );
4568 # Add handler params
4569 $paramMap = $internalParamMap;
4571 $handlerParamMap = $handler->getParamMap();
4572 foreach ( $handlerParamMap as $magic => $paramName ) {
4573 $paramMap[$magic] = array( 'handler', $paramName );
4576 $this->mImageParams
[$handlerClass] = $paramMap;
4577 $this->mImageParamsMagicArray
[$handlerClass] = new MagicWordArray( array_keys( $paramMap ) );
4579 return array( $this->mImageParams
[$handlerClass], $this->mImageParamsMagicArray
[$handlerClass] );
4583 * Parse image options text and use it to make an image
4585 * @param $title Title
4586 * @param $options String
4587 * @param $holders LinkHolderArray
4589 function makeImage( $title, $options, $holders = false ) {
4590 # Check if the options text is of the form "options|alt text"
4592 # * thumbnail make a thumbnail with enlarge-icon and caption, alignment depends on lang
4593 # * left no resizing, just left align. label is used for alt= only
4594 # * right same, but right aligned
4595 # * none same, but not aligned
4596 # * ___px scale to ___ pixels width, no aligning. e.g. use in taxobox
4597 # * center center the image
4598 # * frame Keep original image size, no magnify-button.
4599 # * framed Same as "frame"
4600 # * frameless like 'thumb' but without a frame. Keeps user preferences for width
4601 # * upright reduce width for upright images, rounded to full __0 px
4602 # * border draw a 1px border around the image
4603 # * alt Text for HTML alt attribute (defaults to empty)
4604 # * link Set the target of the image link. Can be external, interwiki, or local
4605 # vertical-align values (no % or length right now):
4615 $parts = StringUtils
::explode( "|", $options );
4616 $sk = $this->mOptions
->getSkin( $this->mTitle
);
4618 # Give extensions a chance to select the file revision for us
4619 $skip = $time = $descQuery = false;
4620 wfRunHooks( 'BeforeParserMakeImageLinkObj', array( &$this, &$title, &$skip, &$time, &$descQuery ) );
4623 return $sk->link( $title );
4627 $file = wfFindFile( $title, array( 'time' => $time ) );
4629 $handler = $file ?
$file->getHandler() : false;
4631 list( $paramMap, $mwArray ) = $this->getImageParams( $handler );
4633 # Process the input parameters
4635 $params = array( 'frame' => array(), 'handler' => array(),
4636 'horizAlign' => array(), 'vertAlign' => array() );
4637 foreach ( $parts as $part ) {
4638 $part = trim( $part );
4639 list( $magicName, $value ) = $mwArray->matchVariableStartToEnd( $part );
4641 if ( isset( $paramMap[$magicName] ) ) {
4642 list( $type, $paramName ) = $paramMap[$magicName];
4644 # Special case; width and height come in one variable together
4645 if ( $type === 'handler' && $paramName === 'width' ) {
4647 # (bug 13500) In both cases (width/height and width only),
4648 # permit trailing "px" for backward compatibility.
4649 if ( preg_match( '/^([0-9]*)x([0-9]*)\s*(?:px)?\s*$/', $value, $m ) ) {
4650 $width = intval( $m[1] );
4651 $height = intval( $m[2] );
4652 if ( $handler->validateParam( 'width', $width ) ) {
4653 $params[$type]['width'] = $width;
4656 if ( $handler->validateParam( 'height', $height ) ) {
4657 $params[$type]['height'] = $height;
4660 } elseif ( preg_match( '/^[0-9]*\s*(?:px)?\s*$/', $value ) ) {
4661 $width = intval( $value );
4662 if ( $handler->validateParam( 'width', $width ) ) {
4663 $params[$type]['width'] = $width;
4666 } # else no validation -- bug 13436
4668 if ( $type === 'handler' ) {
4669 # Validate handler parameter
4670 $validated = $handler->validateParam( $paramName, $value );
4672 # Validate internal parameters
4673 switch( $paramName ) {
4676 # @todo Fixme: possibly check validity here for
4677 # manualthumb? downstream behavior seems odd with
4678 # missing manual thumbs.
4680 $value = $this->stripAltText( $value, $holders );
4683 $chars = self
::EXT_LINK_URL_CLASS
;
4684 $prots = $this->mUrlProtocols
;
4685 if ( $value === '' ) {
4686 $paramName = 'no-link';
4689 } elseif ( preg_match( "/^$prots/", $value ) ) {
4690 if ( preg_match( "/^($prots)$chars+$/", $value, $m ) ) {
4691 $paramName = 'link-url';
4692 $this->mOutput
->addExternalLink( $value );
4696 $linkTitle = Title
::newFromText( $value );
4698 $paramName = 'link-title';
4699 $value = $linkTitle;
4700 $this->mOutput
->addLink( $linkTitle );
4706 # Most other things appear to be empty or numeric...
4707 $validated = ( $value === false ||
is_numeric( trim( $value ) ) );
4712 $params[$type][$paramName] = $value;
4716 if ( !$validated ) {
4721 # Process alignment parameters
4722 if ( $params['horizAlign'] ) {
4723 $params['frame']['align'] = key( $params['horizAlign'] );
4725 if ( $params['vertAlign'] ) {
4726 $params['frame']['valign'] = key( $params['vertAlign'] );
4729 $params['frame']['caption'] = $caption;
4731 # Will the image be presented in a frame, with the caption below?
4732 $imageIsFramed = isset( $params['frame']['frame'] ) ||
4733 isset( $params['frame']['framed'] ) ||
4734 isset( $params['frame']['thumbnail'] ) ||
4735 isset( $params['frame']['manualthumb'] );
4737 # In the old days, [[Image:Foo|text...]] would set alt text. Later it
4738 # came to also set the caption, ordinary text after the image -- which
4739 # makes no sense, because that just repeats the text multiple times in
4740 # screen readers. It *also* came to set the title attribute.
4742 # Now that we have an alt attribute, we should not set the alt text to
4743 # equal the caption: that's worse than useless, it just repeats the
4744 # text. This is the framed/thumbnail case. If there's no caption, we
4745 # use the unnamed parameter for alt text as well, just for the time be-
4746 # ing, if the unnamed param is set and the alt param is not.
4748 # For the future, we need to figure out if we want to tweak this more,
4749 # e.g., introducing a title= parameter for the title; ignoring the un-
4750 # named parameter entirely for images without a caption; adding an ex-
4751 # plicit caption= parameter and preserving the old magic unnamed para-
4753 if ( $imageIsFramed ) { # Framed image
4754 if ( $caption === '' && !isset( $params['frame']['alt'] ) ) {
4755 # No caption or alt text, add the filename as the alt text so
4756 # that screen readers at least get some description of the image
4757 $params['frame']['alt'] = $title->getText();
4759 # Do not set $params['frame']['title'] because tooltips don't make sense
4761 } else { # Inline image
4762 if ( !isset( $params['frame']['alt'] ) ) {
4763 # No alt text, use the "caption" for the alt text
4764 if ( $caption !== '') {
4765 $params['frame']['alt'] = $this->stripAltText( $caption, $holders );
4767 # No caption, fall back to using the filename for the
4769 $params['frame']['alt'] = $title->getText();
4772 # Use the "caption" for the tooltip text
4773 $params['frame']['title'] = $this->stripAltText( $caption, $holders );
4776 wfRunHooks( 'ParserMakeImageParams', array( $title, $file, &$params ) );
4778 # Linker does the rest
4779 $ret = $sk->makeImageLink2( $title, $file, $params['frame'], $params['handler'], $time, $descQuery, $this->mOptions
->getThumbSize() );
4781 # Give the handler a chance to modify the parser object
4783 $handler->parserTransformHook( $this, $file );
4789 protected function stripAltText( $caption, $holders ) {
4790 # Strip bad stuff out of the title (tooltip). We can't just use
4791 # replaceLinkHoldersText() here, because if this function is called
4792 # from replaceInternalLinks2(), mLinkHolders won't be up-to-date.
4794 $tooltip = $holders->replaceText( $caption );
4796 $tooltip = $this->replaceLinkHoldersText( $caption );
4799 # make sure there are no placeholders in thumbnail attributes
4800 # that are later expanded to html- so expand them now and
4802 $tooltip = $this->mStripState
->unstripBoth( $tooltip );
4803 $tooltip = Sanitizer
::stripAllTags( $tooltip );
4809 * Set a flag in the output object indicating that the content is dynamic and
4810 * shouldn't be cached.
4812 function disableCache() {
4813 wfDebug( "Parser output marked as uncacheable.\n" );
4814 $this->mOutput
->setCacheTime( -1 ); // old style, for compatibility
4815 $this->mOutput
->updateCacheExpiry( 0 ); // new style, for consistency
4819 * Callback from the Sanitizer for expanding items found in HTML attribute
4820 * values, so they can be safely tested and escaped.
4822 * @param $text String
4823 * @param $frame PPFrame
4827 function attributeStripCallback( &$text, $frame = false ) {
4828 $text = $this->replaceVariables( $text, $frame );
4829 $text = $this->mStripState
->unstripBoth( $text );
4836 function getTags() {
4837 return array_merge( array_keys( $this->mTransparentTagHooks
), array_keys( $this->mTagHooks
) );
4841 * Break wikitext input into sections, and either pull or replace
4842 * some particular section's text.
4844 * External callers should use the getSection and replaceSection methods.
4846 * @param $text String: Page wikitext
4847 * @param $section String: a section identifier string of the form:
4848 * <flag1> - <flag2> - ... - <section number>
4850 * Currently the only recognised flag is "T", which means the target section number
4851 * was derived during a template inclusion parse, in other words this is a template
4852 * section edit link. If no flags are given, it was an ordinary section edit link.
4853 * This flag is required to avoid a section numbering mismatch when a section is
4854 * enclosed by <includeonly> (bug 6563).
4856 * The section number 0 pulls the text before the first heading; other numbers will
4857 * pull the given section along with its lower-level subsections. If the section is
4858 * not found, $mode=get will return $newtext, and $mode=replace will return $text.
4860 * @param $mode String: one of "get" or "replace"
4861 * @param $newText String: replacement text for section data.
4862 * @return String: for "get", the extracted section text.
4863 * for "replace", the whole page with the section replaced.
4865 private function extractSections( $text, $section, $mode, $newText='' ) {
4867 $this->clearState();
4868 $this->setTitle( $wgTitle ); # not generally used but removes an ugly failure mode
4869 $this->mOptions
= new ParserOptions
;
4870 $this->setOutputType( self
::OT_PLAIN
);
4872 $frame = $this->getPreprocessor()->newFrame();
4874 # Process section extraction flags
4876 $sectionParts = explode( '-', $section );
4877 $sectionIndex = array_pop( $sectionParts );
4878 foreach ( $sectionParts as $part ) {
4879 if ( $part === 'T' ) {
4880 $flags |
= self
::PTD_FOR_INCLUSION
;
4883 # Preprocess the text
4884 $root = $this->preprocessToDom( $text, $flags );
4886 # <h> nodes indicate section breaks
4887 # They can only occur at the top level, so we can find them by iterating the root's children
4888 $node = $root->getFirstChild();
4890 # Find the target section
4891 if ( $sectionIndex == 0 ) {
4892 # Section zero doesn't nest, level=big
4893 $targetLevel = 1000;
4896 if ( $node->getName() === 'h' ) {
4897 $bits = $node->splitHeading();
4898 if ( $bits['i'] == $sectionIndex ) {
4899 $targetLevel = $bits['level'];
4903 if ( $mode === 'replace' ) {
4904 $outText .= $frame->expand( $node, PPFrame
::RECOVER_ORIG
);
4906 $node = $node->getNextSibling();
4912 if ( $mode === 'get' ) {
4919 # Find the end of the section, including nested sections
4921 if ( $node->getName() === 'h' ) {
4922 $bits = $node->splitHeading();
4923 $curLevel = $bits['level'];
4924 if ( $bits['i'] != $sectionIndex && $curLevel <= $targetLevel ) {
4928 if ( $mode === 'get' ) {
4929 $outText .= $frame->expand( $node, PPFrame
::RECOVER_ORIG
);
4931 $node = $node->getNextSibling();
4934 # Write out the remainder (in replace mode only)
4935 if ( $mode === 'replace' ) {
4936 # Output the replacement text
4937 # Add two newlines on -- trailing whitespace in $newText is conventionally
4938 # stripped by the editor, so we need both newlines to restore the paragraph gap
4939 # Only add trailing whitespace if there is newText
4940 if ( $newText != "" ) {
4941 $outText .= $newText . "\n\n";
4945 $outText .= $frame->expand( $node, PPFrame
::RECOVER_ORIG
);
4946 $node = $node->getNextSibling();
4950 if ( is_string( $outText ) ) {
4951 # Re-insert stripped tags
4952 $outText = rtrim( $this->mStripState
->unstripBoth( $outText ) );
4959 * This function returns the text of a section, specified by a number ($section).
4960 * A section is text under a heading like == Heading == or \<h1\>Heading\</h1\>, or
4961 * the first section before any such heading (section 0).
4963 * If a section contains subsections, these are also returned.
4965 * @param $text String: text to look in
4966 * @param $section String: section identifier
4967 * @param $deftext String: default to return if section is not found
4968 * @return string text of the requested section
4970 public function getSection( $text, $section, $deftext='' ) {
4971 return $this->extractSections( $text, $section, "get", $deftext );
4974 public function replaceSection( $oldtext, $section, $text ) {
4975 return $this->extractSections( $oldtext, $section, "replace", $text );
4979 * Get the ID of the revision we are parsing
4981 * @return Mixed: integer or null
4983 function getRevisionId() {
4984 return $this->mRevisionId
;
4988 * Get the timestamp associated with the current revision, adjusted for
4989 * the default server-local timestamp
4991 function getRevisionTimestamp() {
4992 if ( is_null( $this->mRevisionTimestamp
) ) {
4993 wfProfileIn( __METHOD__
);
4995 $dbr = wfGetDB( DB_SLAVE
);
4996 $timestamp = $dbr->selectField( 'revision', 'rev_timestamp',
4997 array( 'rev_id' => $this->mRevisionId
), __METHOD__
);
4999 # Normalize timestamp to internal MW format for timezone processing.
5000 # This has the added side-effect of replacing a null value with
5001 # the current time, which gives us more sensible behavior for
5003 $timestamp = wfTimestamp( TS_MW
, $timestamp );
5005 # The cryptic '' timezone parameter tells to use the site-default
5006 # timezone offset instead of the user settings.
5008 # Since this value will be saved into the parser cache, served
5009 # to other users, and potentially even used inside links and such,
5010 # it needs to be consistent for all visitors.
5011 $this->mRevisionTimestamp
= $wgContLang->userAdjust( $timestamp, '' );
5013 wfProfileOut( __METHOD__
);
5015 return $this->mRevisionTimestamp
;
5019 * Get the name of the user that edited the last revision
5021 * @return String: user name
5023 function getRevisionUser() {
5024 # if this template is subst: the revision id will be blank,
5025 # so just use the current user's name
5026 if ( $this->mRevisionId
) {
5027 $revision = Revision
::newFromId( $this->mRevisionId
);
5028 $revuser = $revision->getUserText();
5031 $revuser = $wgUser->getName();
5037 * Mutator for $mDefaultSort
5039 * @param $sort New value
5041 public function setDefaultSort( $sort ) {
5042 $this->mDefaultSort
= $sort;
5043 $this->mOutput
->setProperty( 'defaultsort', $sort );
5047 * Accessor for $mDefaultSort
5048 * Will use the title/prefixed title if none is set
5052 public function getDefaultSort() {
5053 if ( $this->mDefaultSort
!== false ) {
5054 return $this->mDefaultSort
;
5056 return $this->mTitle
->getCategorySortkey();
5061 * Accessor for $mDefaultSort
5062 * Unlike getDefaultSort(), will return false if none is set
5064 * @return string or false
5066 public function getCustomDefaultSort() {
5067 return $this->mDefaultSort
;
5071 * Try to guess the section anchor name based on a wikitext fragment
5072 * presumably extracted from a heading, for example "Header" from
5075 public function guessSectionNameFromWikiText( $text ) {
5076 # Strip out wikitext links(they break the anchor)
5077 $text = $this->stripSectionName( $text );
5078 $text = Sanitizer
::normalizeSectionNameWhitespace( $text );
5079 return '#' . Sanitizer
::escapeId( $text, 'noninitial' );
5083 * Same as guessSectionNameFromWikiText(), but produces legacy anchors
5084 * instead. For use in redirects, since IE6 interprets Redirect: headers
5085 * as something other than UTF-8 (apparently?), resulting in breakage.
5087 * @param $text String: The section name
5088 * @return string An anchor
5090 public function guessLegacySectionNameFromWikiText( $text ) {
5091 # Strip out wikitext links(they break the anchor)
5092 $text = $this->stripSectionName( $text );
5093 $text = Sanitizer
::normalizeSectionNameWhitespace( $text );
5094 return '#' . Sanitizer
::escapeId( $text, array( 'noninitial', 'legacy' ) );
5098 * Strips a text string of wikitext for use in a section anchor
5100 * Accepts a text string and then removes all wikitext from the
5101 * string and leaves only the resultant text (i.e. the result of
5102 * [[User:WikiSysop|Sysop]] would be "Sysop" and the result of
5103 * [[User:WikiSysop]] would be "User:WikiSysop") - this is intended
5104 * to create valid section anchors by mimicing the output of the
5105 * parser when headings are parsed.
5107 * @param $text String: text string to be stripped of wikitext
5108 * for use in a Section anchor
5109 * @return Filtered text string
5111 public function stripSectionName( $text ) {
5112 # Strip internal link markup
5113 $text = preg_replace( '/\[\[:?([^[|]+)\|([^[]+)\]\]/', '$2', $text );
5114 $text = preg_replace( '/\[\[:?([^[]+)\|?\]\]/', '$1', $text );
5116 # Strip external link markup (FIXME: Not Tolerant to blank link text
5117 # I.E. [http://www.mediawiki.org] will render as [1] or something depending
5118 # on how many empty links there are on the page - need to figure that out.
5119 $text = preg_replace( '/\[(?:' . wfUrlProtocols() . ')([^ ]+?) ([^[]+)\]/', '$2', $text );
5121 # Parse wikitext quotes (italics & bold)
5122 $text = $this->doQuotes( $text );
5125 $text = StringUtils
::delimiterReplace( '<', '>', '', $text );
5130 * strip/replaceVariables/unstrip for preprocessor regression testing
5132 function testSrvus( $text, $title, $options, $outputType = self
::OT_HTML
) {
5133 $this->clearState();
5134 if ( !$title instanceof Title
) {
5135 $title = Title
::newFromText( $title );
5137 $this->mTitle
= $title;
5138 $options->resetUsage();
5139 $this->mOptions
= $options;
5140 $this->setOutputType( $outputType );
5141 $text = $this->replaceVariables( $text );
5142 $text = $this->mStripState
->unstripBoth( $text );
5143 $text = Sanitizer
::removeHTMLtags( $text );
5147 function testPst( $text, $title, $options ) {
5149 if ( !$title instanceof Title
) {
5150 $title = Title
::newFromText( $title );
5152 return $this->preSaveTransform( $text, $title, $wgUser, $options );
5155 function testPreprocess( $text, $title, $options ) {
5156 if ( !$title instanceof Title
) {
5157 $title = Title
::newFromText( $title );
5159 return $this->testSrvus( $text, $title, $options, self
::OT_PREPROCESS
);
5162 function markerSkipCallback( $s, $callback ) {
5165 while ( $i < strlen( $s ) ) {
5166 $markerStart = strpos( $s, $this->mUniqPrefix
, $i );
5167 if ( $markerStart === false ) {
5168 $out .= call_user_func( $callback, substr( $s, $i ) );
5171 $out .= call_user_func( $callback, substr( $s, $i, $markerStart - $i ) );
5172 $markerEnd = strpos( $s, self
::MARKER_SUFFIX
, $markerStart );
5173 if ( $markerEnd === false ) {
5174 $out .= substr( $s, $markerStart );
5177 $markerEnd +
= strlen( self
::MARKER_SUFFIX
);
5178 $out .= substr( $s, $markerStart, $markerEnd - $markerStart );
5186 function serialiseHalfParsedText( $text ) {
5188 $data['text'] = $text;
5190 # First, find all strip markers, and store their
5192 $stripState = new StripState
;
5194 while ( ( $start_pos = strpos( $text, $this->mUniqPrefix
, $pos ) )
5195 && ( $end_pos = strpos( $text, self
::MARKER_SUFFIX
, $pos ) ) )
5197 $end_pos +
= strlen( self
::MARKER_SUFFIX
);
5198 $marker = substr( $text, $start_pos, $end_pos-$start_pos );
5200 if ( !empty( $this->mStripState
->general
->data
[$marker] ) ) {
5201 $replaceArray = $stripState->general
;
5202 $stripText = $this->mStripState
->general
->data
[$marker];
5203 } elseif ( !empty( $this->mStripState
->nowiki
->data
[$marker] ) ) {
5204 $replaceArray = $stripState->nowiki
;
5205 $stripText = $this->mStripState
->nowiki
->data
[$marker];
5207 throw new MWException( "Hanging strip marker: '$marker'." );
5210 $replaceArray->setPair( $marker, $stripText );
5213 $data['stripstate'] = $stripState;
5215 # Now, find all of our links, and store THEIR
5216 # data in an array! :)
5217 $links = array( 'internal' => array(), 'interwiki' => array() );
5221 while ( ( $start_pos = strpos( $text, '<!--LINK ', $pos ) ) ) {
5222 list( $ns, $trail ) = explode( ':', substr( $text, $start_pos +
strlen( '<!--LINK ' ) ), 2 );
5225 if ( empty( $links['internal'][$ns] ) ) {
5226 $links['internal'][$ns] = array();
5229 $key = trim( substr( $trail, 0, strpos( $trail, '-->' ) ) );
5230 $links['internal'][$ns][] = $this->mLinkHolders
->internals
[$ns][$key];
5231 $pos = $start_pos +
strlen( "<!--LINK $ns:$key-->" );
5237 while ( ( $start_pos = strpos( $text, '<!--IWLINK ', $pos ) ) ) {
5238 $data = substr( $text, $start_pos );
5239 $key = trim( substr( $data, 0, strpos( $data, '-->' ) ) );
5240 $links['interwiki'][] = $this->mLinkHolders
->interwiki
[$key];
5241 $pos = $start_pos +
strlen( "<!--IWLINK $key-->" );
5244 $data['linkholder'] = $links;
5251 * @param $data Array
5252 * @param $intPrefix String unique identifying prefix
5255 function unserialiseHalfParsedText( $data, $intPrefix = null ) {
5256 if ( !$intPrefix ) {
5257 $intPrefix = $this->getRandomString();
5260 # First, extract the strip state.
5261 $stripState = $data['stripstate'];
5262 $this->mStripState
->general
->merge( $stripState->general
);
5263 $this->mStripState
->nowiki
->merge( $stripState->nowiki
);
5265 # Now, extract the text, and renumber links
5266 $text = $data['text'];
5267 $links = $data['linkholder'];
5270 foreach ( $links['internal'] as $ns => $nsLinks ) {
5271 foreach ( $nsLinks as $key => $entry ) {
5272 $newKey = $intPrefix . '-' . $key;
5273 $this->mLinkHolders
->internals
[$ns][$newKey] = $entry;
5275 $text = str_replace( "<!--LINK $ns:$key-->", "<!--LINK $ns:$newKey-->", $text );
5280 foreach ( $links['interwiki'] as $key => $entry ) {
5281 $newKey = "$intPrefix-$key";
5282 $this->mLinkHolders
->interwikis
[$newKey] = $entry;
5284 $text = str_replace( "<!--IWLINK $key-->", "<!--IWLINK $newKey-->", $text );
5287 # Should be good to go.
5293 * @todo document, briefly.
5297 var $general, $nowiki;
5299 function __construct() {
5300 $this->general
= new ReplacementArray
;
5301 $this->nowiki
= new ReplacementArray
;
5304 function unstripGeneral( $text ) {
5305 wfProfileIn( __METHOD__
);
5308 $text = $this->general
->replace( $text );
5309 } while ( $text !== $oldText );
5310 wfProfileOut( __METHOD__
);
5314 function unstripNoWiki( $text ) {
5315 wfProfileIn( __METHOD__
);
5318 $text = $this->nowiki
->replace( $text );
5319 } while ( $text !== $oldText );
5320 wfProfileOut( __METHOD__
);
5324 function unstripBoth( $text ) {
5325 wfProfileIn( __METHOD__
);
5328 $text = $this->general
->replace( $text );
5329 $text = $this->nowiki
->replace( $text );
5330 } while ( $text !== $oldText );
5331 wfProfileOut( __METHOD__
);
5337 * @todo document, briefly.
5340 class OnlyIncludeReplacer
{
5343 function replace( $matches ) {
5344 if ( substr( $matches[1], -1 ) === "\n" ) {
5345 $this->output
.= substr( $matches[1], 0, -1 );
5347 $this->output
.= $matches[1];