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
31 * objects: $wgLang, $wgContLang
33 * NOT $wgArticle, $wgUser or $wgTitle. Keep them away!
36 * $wgUseTex*, $wgUseDynamicDates*, $wgInterwikiMagic*,
37 * $wgNamespacesWithSubpages, $wgAllowExternalImages*,
38 * $wgLocaltimezone, $wgAllowSpecialInclusion*,
41 * * only within ParserOptions
49 * Update this version number when the ParserOutput format
50 * changes in an incompatible way, so the parser cache
51 * can automatically discard old data.
53 const VERSION
= '1.6.4';
55 # Flags for Parser::setFunctionHook
56 # Also available as global constants from Defines.php
57 const SFH_NO_HASH
= 1;
58 const SFH_OBJECT_ARGS
= 2;
60 # Constants needed for external link processing
61 # Everything except bracket, space, or control characters
62 const EXT_LINK_URL_CLASS
= '[^][<>"\\x00-\\x20\\x7F]';
63 const EXT_IMAGE_REGEX
= '/^(http:\/\/|https:\/\/)([^][<>"\\x00-\\x20\\x7F]+)
64 \\/([A-Za-z0-9_.,~%\\-+&;#*?!=()@\\x80-\\xFF]+)\\.((?i)gif|png|jpg|jpeg)$/Sx';
66 // State constants for the definition list colon extraction
67 const COLON_STATE_TEXT
= 0;
68 const COLON_STATE_TAG
= 1;
69 const COLON_STATE_TAGSTART
= 2;
70 const COLON_STATE_CLOSETAG
= 3;
71 const COLON_STATE_TAGSLASH
= 4;
72 const COLON_STATE_COMMENT
= 5;
73 const COLON_STATE_COMMENTDASH
= 6;
74 const COLON_STATE_COMMENTDASHDASH
= 7;
76 // Flags for preprocessToDom
77 const PTD_FOR_INCLUSION
= 1;
79 // Allowed values for $this->mOutputType
80 // Parameter to startExternalParse().
83 const OT_PREPROCESS
= 3;
86 // Marker Suffix needs to be accessible staticly.
87 const MARKER_SUFFIX
= "-QINU\x7f";
93 var $mTagHooks, $mTransparentTagHooks, $mFunctionHooks, $mFunctionSynonyms, $mVariables,
94 $mImageParams, $mImageParamsMagicArray, $mStripList, $mMarkerIndex, $mPreprocessor,
95 $mExtLinkBracketedRegex, $mUrlProtocols, $mDefaultStripList, $mVarCache, $mConf;
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
109 # These are variables reset at least once per parse regardless of $clearState
110 var $mOptions, // ParserOptions object
111 $mTitle, // Title context, used for self-link rendering and similar things
112 $mOutputType, // Output type, one of the OT_xxx constants
113 $ot, // Shortcut alias, see setOutputType()
114 $mRevisionId, // ID to display in {{REVISIONID}} tags
115 $mRevisionTimestamp, // The timestamp of the specified revision ID
116 $mRevIdForTs; // The revision ID which was used to fetch the timestamp
125 function __construct( $conf = array() ) {
126 $this->mConf
= $conf;
127 $this->mTagHooks
= array();
128 $this->mTransparentTagHooks
= array();
129 $this->mFunctionHooks
= array();
130 $this->mFunctionSynonyms
= array( 0 => array(), 1 => array() );
131 $this->mDefaultStripList
= $this->mStripList
= array( 'nowiki', 'gallery' );
132 $this->mUrlProtocols
= wfUrlProtocols();
133 $this->mExtLinkBracketedRegex
= '/\[(\b(' . wfUrlProtocols() . ')'.
134 '[^][<>"\\x00-\\x20\\x7F]+) *([^\]\\x0a\\x0d]*?)\]/S';
135 $this->mVarCache
= array();
136 if ( isset( $conf['preprocessorClass'] ) ) {
137 $this->mPreprocessorClass
= $conf['preprocessorClass'];
138 } elseif ( extension_loaded( 'domxml' ) ) {
139 // PECL extension that conflicts with the core DOM extension (bug 13770)
140 wfDebug( "Warning: you have the obsolete domxml extension for PHP. Please remove it!\n" );
141 $this->mPreprocessorClass
= 'Preprocessor_Hash';
142 } elseif ( extension_loaded( 'dom' ) ) {
143 $this->mPreprocessorClass
= 'Preprocessor_DOM';
145 $this->mPreprocessorClass
= 'Preprocessor_Hash';
147 $this->mMarkerIndex
= 0;
148 $this->mFirstCall
= true;
152 * Reduce memory usage to reduce the impact of circular references
154 function __destruct() {
155 if ( isset( $this->mLinkHolders
) ) {
156 $this->mLinkHolders
->__destruct();
158 foreach ( $this as $name => $value ) {
159 unset( $this->$name );
164 * Do various kinds of initialisation on the first call of the parser
166 function firstCallInit() {
167 if ( !$this->mFirstCall
) {
170 $this->mFirstCall
= false;
172 wfProfileIn( __METHOD__
);
174 $this->setHook( 'pre', array( $this, 'renderPreTag' ) );
175 CoreParserFunctions
::register( $this );
176 $this->initialiseVariables();
178 wfRunHooks( 'ParserFirstCallInit', array( &$this ) );
179 wfProfileOut( __METHOD__
);
187 function clearState() {
188 wfProfileIn( __METHOD__
);
189 if ( $this->mFirstCall
) {
190 $this->firstCallInit();
192 $this->mOutput
= new ParserOutput
;
193 $this->mAutonumber
= 0;
194 $this->mLastSection
= '';
195 $this->mDTopen
= false;
196 $this->mIncludeCount
= array();
197 $this->mStripState
= new StripState
;
198 $this->mArgStack
= false;
199 $this->mInPre
= false;
200 $this->mLinkHolders
= new LinkHolderArray( $this );
202 $this->mRevisionTimestamp
= $this->mRevisionId
= null;
205 * Prefix for temporary replacement strings for the multipass parser.
206 * \x07 should never appear in input as it's disallowed in XML.
207 * Using it at the front also gives us a little extra robustness
208 * since it shouldn't match when butted up against identifier-like
211 * Must not consist of all title characters, or else it will change
212 * the behaviour of <nowiki> in a link.
214 #$this->mUniqPrefix = "\x07UNIQ" . Parser::getRandomString();
215 # Changed to \x7f to allow XML double-parsing -- TS
216 $this->mUniqPrefix
= "\x7fUNIQ" . self
::getRandomString();
219 # Clear these on every parse, bug 4549
220 $this->mTplExpandCache
= $this->mTplRedirCache
= $this->mTplDomCache
= array();
222 $this->mShowToc
= true;
223 $this->mForceTocPosition
= false;
224 $this->mIncludeSizes
= array(
228 $this->mPPNodeCount
= 0;
229 $this->mDefaultSort
= false;
230 $this->mHeadings
= array();
231 $this->mDoubleUnderscores
= array();
232 $this->mExpensiveFunctionCount
= 0;
233 $this->mFileCache
= array();
236 if ( isset( $this->mPreprocessor
) && $this->mPreprocessor
->parser
!== $this ) {
237 $this->mPreprocessor
= null;
240 wfRunHooks( 'ParserClearState', array( &$this ) );
241 wfProfileOut( __METHOD__
);
244 function setOutputType( $ot ) {
245 $this->mOutputType
= $ot;
248 'html' => $ot == self
::OT_HTML
,
249 'wiki' => $ot == self
::OT_WIKI
,
250 'pre' => $ot == self
::OT_PREPROCESS
,
255 * Set the context title
257 function setTitle( $t ) {
258 if ( !$t ||
$t instanceof FakeTitle
) {
259 $t = Title
::newFromText( 'NO TITLE' );
261 if ( strval( $t->getFragment() ) !== '' ) {
262 # Strip the fragment to avoid various odd effects
263 $this->mTitle
= clone $t;
264 $this->mTitle
->setFragment( '' );
271 * Accessor for mUniqPrefix.
275 function uniqPrefix() {
276 if( !isset( $this->mUniqPrefix
) ) {
277 // @fixme this is probably *horribly wrong*
278 // LanguageConverter seems to want $wgParser's uniqPrefix, however
279 // if this is called for a parser cache hit, the parser may not
280 // have ever been initialized in the first place.
281 // Not really sure what the heck is supposed to be going on here.
283 //throw new MWException( "Accessing uninitialized mUniqPrefix" );
285 return $this->mUniqPrefix
;
289 * Convert wikitext to HTML
290 * Do not call this function recursively.
292 * @param $text String: text we want to parse
293 * @param $title A title object
294 * @param $options ParserOptions
295 * @param $linestart boolean
296 * @param $clearState boolean
297 * @param $revid Int: number to pass in {{REVISIONID}}
298 * @return ParserOutput a ParserOutput
300 public function parse( $text, Title
$title, ParserOptions
$options, $linestart = true, $clearState = true, $revid = null ) {
302 * First pass--just handle <nowiki> sections, pass the rest off
303 * to internalParse() which does all the real work.
306 global $wgUseTidy, $wgAlwaysUseTidy, $wgContLang;
307 $fname = __METHOD__
.'-' . wfGetCaller();
308 wfProfileIn( __METHOD__
);
309 wfProfileIn( $fname );
315 $this->mOptions
= $options;
316 $this->setTitle( $title );
317 $oldRevisionId = $this->mRevisionId
;
318 $oldRevisionTimestamp = $this->mRevisionTimestamp
;
319 if( $revid !== null ) {
320 $this->mRevisionId
= $revid;
321 $this->mRevisionTimestamp
= null;
323 $this->setOutputType( self
::OT_HTML
);
324 wfRunHooks( 'ParserBeforeStrip', array( &$this, &$text, &$this->mStripState
) );
326 wfRunHooks( 'ParserAfterStrip', array( &$this, &$text, &$this->mStripState
) );
327 $text = $this->internalParse( $text );
328 $text = $this->mStripState
->unstripGeneral( $text );
330 # Clean up special characters, only run once, next-to-last before doBlockLevels
332 # french spaces, last one Guillemet-left
333 # only if there is something before the space
334 '/(.) (?=\\?|:|;|!|%|\\302\\273)/' => '\\1 \\2',
335 # french spaces, Guillemet-right
336 '/(\\302\\253) /' => '\\1 ',
337 '/ (!\s*important)/' => ' \\1', #Beware of CSS magic word !important, bug #11874.
339 $text = preg_replace( array_keys($fixtags), array_values($fixtags), $text );
341 $text = $this->doBlockLevels( $text, $linestart );
343 $this->replaceLinkHolders( $text );
345 # the position of the parserConvert() call should not be changed. it
346 # assumes that the links are all replaced and the only thing left
347 # is the <nowiki> mark.
348 # Side-effects: this calls $this->mOutput->setTitleText()
349 $text = $wgContLang->parserConvert( $text, $this );
351 $text = $this->mStripState
->unstripNoWiki( $text );
353 wfRunHooks( 'ParserBeforeTidy', array( &$this, &$text ) );
355 //!JF Move to its own function
357 $uniq_prefix = $this->mUniqPrefix
;
359 $elements = array_keys( $this->mTransparentTagHooks
);
360 $text = self
::extractTagsAndParams( $elements, $text, $matches, $uniq_prefix );
362 foreach( $matches as $marker => $data ) {
363 list( $element, $content, $params, $tag ) = $data;
364 $tagName = strtolower( $element );
365 if( isset( $this->mTransparentTagHooks
[$tagName] ) ) {
366 $output = call_user_func_array( $this->mTransparentTagHooks
[$tagName],
367 array( $content, $params, $this ) );
371 $this->mStripState
->general
->setPair( $marker, $output );
373 $text = $this->mStripState
->unstripGeneral( $text );
375 $text = Sanitizer
::normalizeCharReferences( $text );
377 if (($wgUseTidy and $this->mOptions
->mTidy
) or $wgAlwaysUseTidy) {
378 $text = self
::tidy($text);
380 # attempt to sanitize at least some nesting problems
381 # (bug #2702 and quite a few others)
383 # ''Something [http://www.cool.com cool''] -->
384 # <i>Something</i><a href="http://www.cool.com"..><i>cool></i></a>
385 '/(<([bi])>)(<([bi])>)?([^<]*)(<\/?a[^<]*>)([^<]*)(<\/\\4>)?(<\/\\2>)/' =>
386 '\\1\\3\\5\\8\\9\\6\\1\\3\\7\\8\\9',
387 # fix up an anchor inside another anchor, only
388 # at least for a single single nested link (bug 3695)
389 '/(<a[^>]+>)([^<]*)(<a[^>]+>[^<]*)<\/a>(.*)<\/a>/' =>
390 '\\1\\2</a>\\3</a>\\1\\4</a>',
391 # fix div inside inline elements- doBlockLevels won't wrap a line which
392 # contains a div, so fix it up here; replace
393 # div with escaped text
394 '/(<([aib]) [^>]+>)([^<]*)(<div([^>]*)>)(.*)(<\/div>)([^<]*)(<\/\\2>)/' =>
395 '\\1\\3<div\\5>\\6</div>\\8\\9',
396 # remove empty italic or bold tag pairs, some
397 # introduced by rules above
398 '/<([bi])><\/\\1>/' => '',
401 $text = preg_replace(
402 array_keys( $tidyregs ),
403 array_values( $tidyregs ),
406 global $wgExpensiveParserFunctionLimit;
407 if ( $this->mExpensiveFunctionCount
> $wgExpensiveParserFunctionLimit ) {
408 $this->limitationWarn( 'expensive-parserfunction', $this->mExpensiveFunctionCount
, $wgExpensiveParserFunctionLimit );
411 wfRunHooks( 'ParserAfterTidy', array( &$this, &$text ) );
413 # Information on include size limits, for the benefit of users who try to skirt them
414 if ( $this->mOptions
->getEnableLimitReport() ) {
415 global $wgExpensiveParserFunctionLimit;
416 $max = $this->mOptions
->getMaxIncludeSize();
417 $PFreport = "Expensive parser function count: {$this->mExpensiveFunctionCount}/$wgExpensiveParserFunctionLimit\n";
419 "NewPP limit report\n" .
420 "Preprocessor node count: {$this->mPPNodeCount}/{$this->mOptions->mMaxPPNodeCount}\n" .
421 "Post-expand include size: {$this->mIncludeSizes['post-expand']}/$max bytes\n" .
422 "Template argument size: {$this->mIncludeSizes['arg']}/$max bytes\n".
424 wfRunHooks( 'ParserLimitReport', array( $this, &$limitReport ) );
425 $text .= "\n<!-- \n$limitReport-->\n";
427 $this->mOutput
->setText( $text );
428 $this->mRevisionId
= $oldRevisionId;
429 $this->mRevisionTimestamp
= $oldRevisionTimestamp;
430 wfProfileOut( $fname );
431 wfProfileOut( __METHOD__
);
433 return $this->mOutput
;
437 * Recursive parser entry point that can be called from an extension tag
440 function recursiveTagParse( $text ) {
441 wfProfileIn( __METHOD__
);
442 wfRunHooks( 'ParserBeforeStrip', array( &$this, &$text, &$this->mStripState
) );
443 wfRunHooks( 'ParserAfterStrip', array( &$this, &$text, &$this->mStripState
) );
444 $text = $this->internalParse( $text );
445 wfProfileOut( __METHOD__
);
450 * Expand templates and variables in the text, producing valid, static wikitext.
451 * Also removes comments.
453 function preprocess( $text, $title, $options, $revid = null ) {
454 wfProfileIn( __METHOD__
);
456 $this->setOutputType( self
::OT_PREPROCESS
);
457 $this->mOptions
= $options;
458 $this->setTitle( $title );
459 if( $revid !== null ) {
460 $this->mRevisionId
= $revid;
462 wfRunHooks( 'ParserBeforeStrip', array( &$this, &$text, &$this->mStripState
) );
463 wfRunHooks( 'ParserAfterStrip', array( &$this, &$text, &$this->mStripState
) );
464 $text = $this->replaceVariables( $text );
465 $text = $this->mStripState
->unstripBoth( $text );
466 wfProfileOut( __METHOD__
);
471 * Get a random string
476 function getRandomString() {
477 return dechex(mt_rand(0, 0x7fffffff)) . dechex(mt_rand(0, 0x7fffffff));
480 function &getTitle() { return $this->mTitle
; }
481 function getOptions() { return $this->mOptions
; }
482 function getRevisionId() { return $this->mRevisionId
; }
483 function getOutput() { return $this->mOutput
; }
484 function nextLinkID() { return $this->mLinkID++
; }
486 function getFunctionLang() {
487 global $wgLang, $wgContLang;
489 $target = $this->mOptions
->getTargetLanguage();
490 if ( $target !== null ) {
493 return $this->mOptions
->getInterfaceMessage() ?
$wgLang : $wgContLang;
498 * Get a preprocessor object
500 function getPreprocessor() {
501 if ( !isset( $this->mPreprocessor
) ) {
502 $class = $this->mPreprocessorClass
;
503 $this->mPreprocessor
= new $class( $this );
505 return $this->mPreprocessor
;
509 * Replaces all occurrences of HTML-style comments and the given tags
510 * in the text with a random marker and returns the next text. The output
511 * parameter $matches will be an associative array filled with data in
513 * 'UNIQ-xxxxx' => array(
516 * array( 'param' => 'x' ),
517 * '<element param="x">tag content</element>' ) )
519 * @param $elements list of element names. Comments are always extracted.
520 * @param $text Source text string.
521 * @param $uniq_prefix
526 function extractTagsAndParams($elements, $text, &$matches, $uniq_prefix = ''){
531 $taglist = implode( '|', $elements );
532 $start = "/<($taglist)(\\s+[^>]*?|\\s*?)(\/?>)|<(!--)/i";
534 while ( '' != $text ) {
535 $p = preg_split( $start, $text, 2, PREG_SPLIT_DELIM_CAPTURE
);
537 if( count( $p ) < 5 ) {
540 if( count( $p ) > 5 ) {
554 $marker = "$uniq_prefix-$element-" . sprintf('%08X', $n++
) . self
::MARKER_SUFFIX
;
555 $stripped .= $marker;
557 if ( $close === '/>' ) {
558 // Empty element tag, <tag />
563 if( $element === '!--' ) {
566 $end = "/(<\\/$element\\s*>)/i";
568 $q = preg_split( $end, $inside, 2, PREG_SPLIT_DELIM_CAPTURE
);
570 if( count( $q ) < 3 ) {
571 # No end tag -- let it run out to the end of the text.
580 $matches[$marker] = array( $element,
582 Sanitizer
::decodeTagAttributes( $attributes ),
583 "<$element$attributes$close$content$tail" );
589 * Get a list of strippable XML-like elements
591 function getStripList() {
593 $elements = $this->mStripList
;
595 $elements[] = 'html';
597 if( $this->mOptions
->getUseTeX() ) {
598 $elements[] = 'math';
604 * @deprecated use replaceVariables
606 function strip( $text, $state, $stripcomments = false , $dontstrip = array () ) {
611 * Restores pre, math, and other extensions removed by strip()
613 * always call unstripNoWiki() after this one
615 * @deprecated use $this->mStripState->unstrip()
617 function unstrip( $text, $state ) {
618 return $state->unstripGeneral( $text );
622 * Always call this after unstrip() to preserve the order
625 * @deprecated use $this->mStripState->unstrip()
627 function unstripNoWiki( $text, $state ) {
628 return $state->unstripNoWiki( $text );
632 * @deprecated use $this->mStripState->unstripBoth()
634 function unstripForHTML( $text ) {
635 return $this->mStripState
->unstripBoth( $text );
639 * Add an item to the strip state
640 * Returns the unique tag which must be inserted into the stripped text
641 * The tag will be replaced with the original text in unstrip()
645 function insertStripItem( $text ) {
646 $rnd = "{$this->mUniqPrefix}-item-{$this->mMarkerIndex}-" . self
::MARKER_SUFFIX
;
647 $this->mMarkerIndex++
;
648 $this->mStripState
->general
->setPair( $rnd, $text );
653 * Interface with html tidy, used if $wgUseTidy = true.
654 * If tidy isn't able to correct the markup, the original will be
655 * returned in all its glory with a warning comment appended.
657 * Either the external tidy program or the in-process tidy extension
658 * will be used depending on availability. Override the default
659 * $wgTidyInternal setting to disable the internal if it's not working.
661 * @param string $text Hideous HTML input
662 * @return string Corrected HTML output
666 function tidy( $text ) {
667 global $wgTidyInternal;
669 $wrappedtext = '<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN"'.
670 ' "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"><html>'.
671 '<head><title>test</title></head><body>'.$text.'</body></html>';
673 # Tidy is known to clobber tabs; convert 'em to entities
674 $wrappedtext = str_replace("\t", '	', $wrappedtext);
676 if( $wgTidyInternal ) {
677 $correctedtext = self
::internalTidy( $wrappedtext );
679 $correctedtext = self
::externalTidy( $wrappedtext );
681 if( is_null( $correctedtext ) ) {
682 wfDebug( "Tidy error detected!\n" );
683 return $text . "\n<!-- Tidy found serious XHTML errors -->\n";
686 # Convert the tabs back from entities
687 $correctedtext = str_replace('	', "\t", $correctedtext);
689 return $correctedtext;
693 * Spawn an external HTML tidy process and get corrected markup back from it.
698 function externalTidy( $text ) {
699 global $wgTidyConf, $wgTidyBin, $wgTidyOpts;
700 wfProfileIn( __METHOD__
);
705 $descriptorspec = array(
706 0 => array('pipe', 'r'),
707 1 => array('pipe', 'w'),
708 2 => array('file', wfGetNull(), 'a')
711 if( function_exists('proc_open') ) {
712 $process = proc_open("$wgTidyBin -config $wgTidyConf $wgTidyOpts$opts", $descriptorspec, $pipes);
713 if (is_resource($process)) {
714 // Theoretically, this style of communication could cause a deadlock
715 // here. If the stdout buffer fills up, then writes to stdin could
716 // block. This doesn't appear to happen with tidy, because tidy only
717 // writes to stdout after it's finished reading from stdin. Search
718 // for tidyParseStdin and tidySaveStdout in console/tidy.c
719 fwrite($pipes[0], $text);
721 while (!feof($pipes[1])) {
722 $cleansource .= fgets($pipes[1], 1024);
725 proc_close($process);
729 wfProfileOut( __METHOD__
);
731 if( $cleansource == '' && $text != '') {
732 // Some kind of error happened, so we couldn't get the corrected text.
733 // Just give up; we'll use the source text and append a warning.
741 * Use the HTML tidy PECL extension to use the tidy library in-process,
742 * saving the overhead of spawning a new process.
744 * 'pear install tidy' should be able to compile the extension module.
749 function internalTidy( $text ) {
750 global $wgTidyConf, $IP, $wgDebugTidy;
751 wfProfileIn( __METHOD__
);
754 $tidy->parseString( $text, $wgTidyConf, 'utf8' );
755 $tidy->cleanRepair();
756 if( $tidy->getStatus() == 2 ) {
757 // 2 is magic number for fatal error
758 // http://www.php.net/manual/en/function.tidy-get-status.php
761 $cleansource = tidy_get_output( $tidy );
763 if ( $wgDebugTidy && $tidy->getStatus() > 0 ) {
764 $cleansource .= "<!--\nTidy reports:\n" .
765 str_replace( '-->', '-->', $tidy->errorBuffer
) .
769 wfProfileOut( __METHOD__
);
774 * parse the wiki syntax used to render tables
778 function doTableStuff ( $text ) {
779 wfProfileIn( __METHOD__
);
781 $lines = StringUtils
::explode( "\n", $text );
783 $td_history = array (); // Is currently a td tag open?
784 $last_tag_history = array (); // Save history of last lag activated (td, th or caption)
785 $tr_history = array (); // Is currently a tr tag open?
786 $tr_attributes = array (); // history of tr attributes
787 $has_opened_tr = array(); // Did this table open a <tr> element?
788 $indent_level = 0; // indent level of the table
790 foreach ( $lines as $outLine ) {
791 $line = trim( $outLine );
793 if( $line == '' ) { // empty line, go to next line
794 $out .= $outLine."\n";
797 $first_character = $line[0];
800 if ( preg_match( '/^(:*)\{\|(.*)$/', $line , $matches ) ) {
801 // First check if we are starting a new table
802 $indent_level = strlen( $matches[1] );
804 $attributes = $this->mStripState
->unstripBoth( $matches[2] );
805 $attributes = Sanitizer
::fixTagAttributes ( $attributes , 'table' );
807 $outLine = str_repeat( '<dl><dd>' , $indent_level ) . "<table{$attributes}>";
808 array_push ( $td_history , false );
809 array_push ( $last_tag_history , '' );
810 array_push ( $tr_history , false );
811 array_push ( $tr_attributes , '' );
812 array_push ( $has_opened_tr , false );
813 } else if ( count ( $td_history ) == 0 ) {
814 // Don't do any of the following
815 $out .= $outLine."\n";
817 } else if ( substr ( $line , 0 , 2 ) === '|}' ) {
818 // We are ending a table
819 $line = '</table>' . substr ( $line , 2 );
820 $last_tag = array_pop ( $last_tag_history );
822 if ( !array_pop ( $has_opened_tr ) ) {
823 $line = "<tr><td></td></tr>{$line}";
826 if ( array_pop ( $tr_history ) ) {
827 $line = "</tr>{$line}";
830 if ( array_pop ( $td_history ) ) {
831 $line = "</{$last_tag}>{$line}";
833 array_pop ( $tr_attributes );
834 $outLine = $line . str_repeat( '</dd></dl>' , $indent_level );
835 } else if ( substr ( $line , 0 , 2 ) === '|-' ) {
836 // Now we have a table row
837 $line = preg_replace( '#^\|-+#', '', $line );
839 // Whats after the tag is now only attributes
840 $attributes = $this->mStripState
->unstripBoth( $line );
841 $attributes = Sanitizer
::fixTagAttributes ( $attributes , 'tr' );
842 array_pop ( $tr_attributes );
843 array_push ( $tr_attributes , $attributes );
846 $last_tag = array_pop ( $last_tag_history );
847 array_pop ( $has_opened_tr );
848 array_push ( $has_opened_tr , true );
850 if ( array_pop ( $tr_history ) ) {
854 if ( array_pop ( $td_history ) ) {
855 $line = "</{$last_tag}>{$line}";
859 array_push ( $tr_history , false );
860 array_push ( $td_history , false );
861 array_push ( $last_tag_history , '' );
863 else if ( $first_character === '|' ||
$first_character === '!' ||
substr ( $line , 0 , 2 ) === '|+' ) {
864 // This might be cell elements, td, th or captions
865 if ( substr ( $line , 0 , 2 ) === '|+' ) {
866 $first_character = '+';
867 $line = substr ( $line , 1 );
870 $line = substr ( $line , 1 );
872 if ( $first_character === '!' ) {
873 $line = str_replace ( '!!' , '||' , $line );
876 // Split up multiple cells on the same line.
877 // FIXME : This can result in improper nesting of tags processed
878 // by earlier parser steps, but should avoid splitting up eg
879 // attribute values containing literal "||".
880 $cells = StringUtils
::explodeMarkup( '||' , $line );
884 // Loop through each table cell
885 foreach ( $cells as $cell )
888 if ( $first_character !== '+' )
890 $tr_after = array_pop ( $tr_attributes );
891 if ( !array_pop ( $tr_history ) ) {
892 $previous = "<tr{$tr_after}>\n";
894 array_push ( $tr_history , true );
895 array_push ( $tr_attributes , '' );
896 array_pop ( $has_opened_tr );
897 array_push ( $has_opened_tr , true );
900 $last_tag = array_pop ( $last_tag_history );
902 if ( array_pop ( $td_history ) ) {
903 $previous = "</{$last_tag}>{$previous}";
906 if ( $first_character === '|' ) {
908 } else if ( $first_character === '!' ) {
910 } else if ( $first_character === '+' ) {
911 $last_tag = 'caption';
916 array_push ( $last_tag_history , $last_tag );
918 // A cell could contain both parameters and data
919 $cell_data = explode ( '|' , $cell , 2 );
921 // Bug 553: Note that a '|' inside an invalid link should not
922 // be mistaken as delimiting cell parameters
923 if ( strpos( $cell_data[0], '[[' ) !== false ) {
924 $cell = "{$previous}<{$last_tag}>{$cell}";
925 } else if ( count ( $cell_data ) == 1 )
926 $cell = "{$previous}<{$last_tag}>{$cell_data[0]}";
928 $attributes = $this->mStripState
->unstripBoth( $cell_data[0] );
929 $attributes = Sanitizer
::fixTagAttributes( $attributes , $last_tag );
930 $cell = "{$previous}<{$last_tag}{$attributes}>{$cell_data[1]}";
934 array_push ( $td_history , true );
937 $out .= $outLine . "\n";
940 // Closing open td, tr && table
941 while ( count ( $td_history ) > 0 )
943 if ( array_pop ( $td_history ) ) {
946 if ( array_pop ( $tr_history ) ) {
949 if ( !array_pop ( $has_opened_tr ) ) {
950 $out .= "<tr><td></td></tr>\n" ;
953 $out .= "</table>\n";
956 // Remove trailing line-ending (b/c)
957 if ( substr( $out, -1 ) === "\n" ) {
958 $out = substr( $out, 0, -1 );
961 // special case: don't return empty table
962 if( $out === "<table>\n<tr><td></td></tr>\n</table>" ) {
966 wfProfileOut( __METHOD__
);
972 * Helper function for parse() that transforms wiki markup into
973 * HTML. Only called for $mOutputType == self::OT_HTML.
977 function internalParse( $text ) {
979 wfProfileIn( __METHOD__
);
981 # Hook to suspend the parser in this state
982 if ( !wfRunHooks( 'ParserBeforeInternalParse', array( &$this, &$text, &$this->mStripState
) ) ) {
983 wfProfileOut( __METHOD__
);
987 $text = $this->replaceVariables( $text );
988 $text = Sanitizer
::removeHTMLtags( $text, array( &$this, 'attributeStripCallback' ), false, array_keys( $this->mTransparentTagHooks
) );
989 wfRunHooks( 'InternalParseBeforeLinks', array( &$this, &$text, &$this->mStripState
) );
991 // Tables need to come after variable replacement for things to work
992 // properly; putting them before other transformations should keep
993 // exciting things like link expansions from showing up in surprising
995 $text = $this->doTableStuff( $text );
997 $text = preg_replace( '/(^|\n)-----*/', '\\1<hr />', $text );
999 $text = $this->doDoubleUnderscore( $text );
1000 $text = $this->doHeadings( $text );
1001 if($this->mOptions
->getUseDynamicDates()) {
1002 $df = DateFormatter
::getInstance();
1003 $text = $df->reformat( $this->mOptions
->getDateFormat(), $text );
1005 $text = $this->doAllQuotes( $text );
1006 $text = $this->replaceInternalLinks( $text );
1007 $text = $this->replaceExternalLinks( $text );
1009 # replaceInternalLinks may sometimes leave behind
1010 # absolute URLs, which have to be masked to hide them from replaceExternalLinks
1011 $text = str_replace($this->mUniqPrefix
."NOPARSE", "", $text);
1013 $text = $this->doMagicLinks( $text );
1014 $text = $this->formatHeadings( $text, $isMain );
1016 wfProfileOut( __METHOD__
);
1021 * Replace special strings like "ISBN xxx" and "RFC xxx" with
1022 * magic external links.
1027 function doMagicLinks( $text ) {
1028 wfProfileIn( __METHOD__
);
1029 $prots = $this->mUrlProtocols
;
1030 $urlChar = self
::EXT_LINK_URL_CLASS
;
1031 $text = preg_replace_callback(
1033 (<a.*?</a>) | # m[1]: Skip link text
1034 (<.*?>) | # m[2]: Skip stuff inside HTML elements' . "
1035 (\\b(?:$prots)$urlChar+) | # m[3]: Free external links" . '
1036 (?:RFC|PMID)\s+([0-9]+) | # m[4]: RFC or PMID, capture number
1037 ISBN\s+(\b # m[5]: ISBN, capture number
1038 (?: 97[89] [\ \-]? )? # optional 13-digit ISBN prefix
1039 (?: [0-9] [\ \-]? ){9} # 9 digits with opt. delimiters
1040 [0-9Xx] # check digit
1042 )!x', array( &$this, 'magicLinkCallback' ), $text );
1043 wfProfileOut( __METHOD__
);
1047 function magicLinkCallback( $m ) {
1048 if ( isset( $m[1] ) && strval( $m[1] ) !== '' ) {
1051 } elseif ( isset( $m[2] ) && strval( $m[2] ) !== '' ) {
1054 } elseif ( isset( $m[3] ) && strval( $m[3] ) !== '' ) {
1055 # Free external link
1056 return $this->makeFreeExternalLink( $m[0] );
1057 } elseif ( isset( $m[4] ) && strval( $m[4] ) !== '' ) {
1059 if ( substr( $m[0], 0, 3 ) === 'RFC' ) {
1063 } elseif ( substr( $m[0], 0, 4 ) === 'PMID' ) {
1065 $urlmsg = 'pubmedurl';
1068 throw new MWException( __METHOD__
.': unrecognised match type "' .
1069 substr($m[0], 0, 20 ) . '"' );
1071 $url = wfMsg( $urlmsg, $id);
1072 $sk = $this->mOptions
->getSkin();
1073 $la = $sk->getExternalLinkAttributes( $url, $keyword.$id );
1074 return "<a href=\"{$url}\"{$la}>{$keyword} {$id}</a>";
1075 } elseif ( isset( $m[5] ) && strval( $m[5] ) !== '' ) {
1078 $num = strtr( $isbn, array(
1083 $titleObj = SpecialPage
::getTitleFor( 'Booksources', $num );
1085 $titleObj->escapeLocalUrl() .
1086 "\" class=\"internal\">ISBN $isbn</a>";
1093 * Make a free external link, given a user-supplied URL
1097 function makeFreeExternalLink( $url ) {
1099 wfProfileIn( __METHOD__
);
1101 $sk = $this->mOptions
->getSkin();
1104 # The characters '<' and '>' (which were escaped by
1105 # removeHTMLtags()) should not be included in
1106 # URLs, per RFC 2396.
1108 if (preg_match('/&(lt|gt);/', $url, $m2, PREG_OFFSET_CAPTURE
)) {
1109 $trail = substr($url, $m2[0][1]) . $trail;
1110 $url = substr($url, 0, $m2[0][1]);
1113 # Move trailing punctuation to $trail
1115 # If there is no left bracket, then consider right brackets fair game too
1116 if ( strpos( $url, '(' ) === false ) {
1120 $numSepChars = strspn( strrev( $url ), $sep );
1121 if ( $numSepChars ) {
1122 $trail = substr( $url, -$numSepChars ) . $trail;
1123 $url = substr( $url, 0, -$numSepChars );
1126 $url = Sanitizer
::cleanUrl( $url );
1128 # Is this an external image?
1129 $text = $this->maybeMakeExternalImage( $url );
1130 if ( $text === false ) {
1131 # Not an image, make a link
1132 $text = $sk->makeExternalLink( $url, $wgContLang->markNoConversion($url), true, 'free',
1133 $this->getExternalLinkAttribs( $url ) );
1134 # Register it in the output object...
1135 # Replace unnecessary URL escape codes with their equivalent characters
1136 $pasteurized = self
::replaceUnusualEscapes( $url );
1137 $this->mOutput
->addExternalLink( $pasteurized );
1139 wfProfileOut( __METHOD__
);
1140 return $text . $trail;
1145 * Parse headers and return html
1149 function doHeadings( $text ) {
1150 wfProfileIn( __METHOD__
);
1151 for ( $i = 6; $i >= 1; --$i ) {
1152 $h = str_repeat( '=', $i );
1153 $text = preg_replace( "/^$h(.+)$h\\s*$/m",
1154 "<h$i>\\1</h$i>", $text );
1156 wfProfileOut( __METHOD__
);
1161 * Replace single quotes with HTML markup
1163 * @return string the altered text
1165 function doAllQuotes( $text ) {
1166 wfProfileIn( __METHOD__
);
1168 $lines = StringUtils
::explode( "\n", $text );
1169 foreach ( $lines as $line ) {
1170 $outtext .= $this->doQuotes( $line ) . "\n";
1172 $outtext = substr($outtext, 0,-1);
1173 wfProfileOut( __METHOD__
);
1178 * Helper function for doAllQuotes()
1180 public function doQuotes( $text ) {
1181 $arr = preg_split( "/(''+)/", $text, -1, PREG_SPLIT_DELIM_CAPTURE
);
1182 if ( count( $arr ) == 1 )
1186 # First, do some preliminary work. This may shift some apostrophes from
1187 # being mark-up to being text. It also counts the number of occurrences
1188 # of bold and italics mark-ups.
1192 foreach ( $arr as $r )
1194 if ( ( $i %
2 ) == 1 )
1196 # If there are ever four apostrophes, assume the first is supposed to
1197 # be text, and the remaining three constitute mark-up for bold text.
1198 if ( strlen( $arr[$i] ) == 4 )
1203 # If there are more than 5 apostrophes in a row, assume they're all
1204 # text except for the last 5.
1205 else if ( strlen( $arr[$i] ) > 5 )
1207 $arr[$i-1] .= str_repeat( "'", strlen( $arr[$i] ) - 5 );
1210 # Count the number of occurrences of bold and italics mark-ups.
1211 # We are not counting sequences of five apostrophes.
1212 if ( strlen( $arr[$i] ) == 2 ) { $numitalics++
; }
1213 else if ( strlen( $arr[$i] ) == 3 ) { $numbold++
; }
1214 else if ( strlen( $arr[$i] ) == 5 ) { $numitalics++
; $numbold++
; }
1219 # If there is an odd number of both bold and italics, it is likely
1220 # that one of the bold ones was meant to be an apostrophe followed
1221 # by italics. Which one we cannot know for certain, but it is more
1222 # likely to be one that has a single-letter word before it.
1223 if ( ( $numbold %
2 == 1 ) && ( $numitalics %
2 == 1 ) )
1226 $firstsingleletterword = -1;
1227 $firstmultiletterword = -1;
1229 foreach ( $arr as $r )
1231 if ( ( $i %
2 == 1 ) and ( strlen( $r ) == 3 ) )
1233 $x1 = substr ($arr[$i-1], -1);
1234 $x2 = substr ($arr[$i-1], -2, 1);
1236 if ($firstspace == -1) $firstspace = $i;
1237 } else if ($x2 === ' ') {
1238 if ($firstsingleletterword == -1) $firstsingleletterword = $i;
1240 if ($firstmultiletterword == -1) $firstmultiletterword = $i;
1246 # If there is a single-letter word, use it!
1247 if ($firstsingleletterword > -1)
1249 $arr [ $firstsingleletterword ] = "''";
1250 $arr [ $firstsingleletterword-1 ] .= "'";
1252 # If not, but there's a multi-letter word, use that one.
1253 else if ($firstmultiletterword > -1)
1255 $arr [ $firstmultiletterword ] = "''";
1256 $arr [ $firstmultiletterword-1 ] .= "'";
1258 # ... otherwise use the first one that has neither.
1259 # (notice that it is possible for all three to be -1 if, for example,
1260 # there is only one pentuple-apostrophe in the line)
1261 else if ($firstspace > -1)
1263 $arr [ $firstspace ] = "''";
1264 $arr [ $firstspace-1 ] .= "'";
1268 # Now let's actually convert our apostrophic mush to HTML!
1273 foreach ($arr as $r)
1277 if ($state === 'both')
1284 if (strlen ($r) == 2)
1287 { $output .= '</i>'; $state = ''; }
1288 else if ($state === 'bi')
1289 { $output .= '</i>'; $state = 'b'; }
1290 else if ($state === 'ib')
1291 { $output .= '</b></i><b>'; $state = 'b'; }
1292 else if ($state === 'both')
1293 { $output .= '<b><i>'.$buffer.'</i>'; $state = 'b'; }
1294 else # $state can be 'b' or ''
1295 { $output .= '<i>'; $state .= 'i'; }
1297 else if (strlen ($r) == 3)
1300 { $output .= '</b>'; $state = ''; }
1301 else if ($state === 'bi')
1302 { $output .= '</i></b><i>'; $state = 'i'; }
1303 else if ($state === 'ib')
1304 { $output .= '</b>'; $state = 'i'; }
1305 else if ($state === 'both')
1306 { $output .= '<i><b>'.$buffer.'</b>'; $state = 'i'; }
1307 else # $state can be 'i' or ''
1308 { $output .= '<b>'; $state .= 'b'; }
1310 else if (strlen ($r) == 5)
1313 { $output .= '</b><i>'; $state = 'i'; }
1314 else if ($state === 'i')
1315 { $output .= '</i><b>'; $state = 'b'; }
1316 else if ($state === 'bi')
1317 { $output .= '</i></b>'; $state = ''; }
1318 else if ($state === 'ib')
1319 { $output .= '</b></i>'; $state = ''; }
1320 else if ($state === 'both')
1321 { $output .= '<i><b>'.$buffer.'</b></i>'; $state = ''; }
1322 else # ($state == '')
1323 { $buffer = ''; $state = 'both'; }
1328 # Now close all remaining tags. Notice that the order is important.
1329 if ($state === 'b' ||
$state === 'ib')
1331 if ($state === 'i' ||
$state === 'bi' ||
$state === 'ib')
1333 if ($state === 'bi')
1335 # There might be lonely ''''', so make sure we have a buffer
1336 if ($state === 'both' && $buffer)
1337 $output .= '<b><i>'.$buffer.'</i></b>';
1343 * Replace external links (REL)
1345 * Note: this is all very hackish and the order of execution matters a lot.
1346 * Make sure to run maintenance/parserTests.php if you change this code.
1350 function replaceExternalLinks( $text ) {
1352 wfProfileIn( __METHOD__
);
1354 $sk = $this->mOptions
->getSkin();
1356 $bits = preg_split( $this->mExtLinkBracketedRegex
, $text, -1, PREG_SPLIT_DELIM_CAPTURE
);
1357 $s = array_shift( $bits );
1360 while ( $i<count( $bits ) ) {
1362 $protocol = $bits[$i++
];
1363 $text = $bits[$i++
];
1364 $trail = $bits[$i++
];
1366 # The characters '<' and '>' (which were escaped by
1367 # removeHTMLtags()) should not be included in
1368 # URLs, per RFC 2396.
1370 if (preg_match('/&(lt|gt);/', $url, $m2, PREG_OFFSET_CAPTURE
)) {
1371 $text = substr($url, $m2[0][1]) . ' ' . $text;
1372 $url = substr($url, 0, $m2[0][1]);
1375 # If the link text is an image URL, replace it with an <img> tag
1376 # This happened by accident in the original parser, but some people used it extensively
1377 $img = $this->maybeMakeExternalImage( $text );
1378 if ( $img !== false ) {
1384 # Set linktype for CSS - if URL==text, link is essentially free
1385 $linktype = ($text === $url) ?
'free' : 'text';
1387 # No link text, e.g. [http://domain.tld/some.link]
1388 if ( $text == '' ) {
1389 # Autonumber if allowed. See bug #5918
1390 if ( strpos( wfUrlProtocols(), substr($protocol, 0, strpos($protocol, ':')) ) !== false ) {
1391 $langObj = $this->getFunctionLang();
1392 $text = '[' . $langObj->formatNum( ++
$this->mAutonumber
) . ']';
1393 $linktype = 'autonumber';
1395 # Otherwise just use the URL
1396 $text = htmlspecialchars( $url );
1400 # Have link text, e.g. [http://domain.tld/some.link text]s
1402 list( $dtrail, $trail ) = Linker
::splitTrail( $trail );
1405 $text = $wgContLang->markNoConversion($text);
1407 $url = Sanitizer
::cleanUrl( $url );
1409 # Use the encoded URL
1410 # This means that users can paste URLs directly into the text
1411 # Funny characters like ö aren't valid in URLs anyway
1412 # This was changed in August 2004
1413 $s .= $sk->makeExternalLink( $url, $text, false, $linktype,
1414 $this->getExternalLinkAttribs( $url ) ) . $dtrail . $trail;
1416 # Register link in the output object.
1417 # Replace unnecessary URL escape codes with the referenced character
1418 # This prevents spammers from hiding links from the filters
1419 $pasteurized = self
::replaceUnusualEscapes( $url );
1420 $this->mOutput
->addExternalLink( $pasteurized );
1423 wfProfileOut( __METHOD__
);
1428 * Get an associative array of additional HTML attributes appropriate for a
1429 * particular external link. This currently may include rel => nofollow
1430 * (depending on configuration, namespace, and the URL's domain) and/or a
1431 * target attribute (depending on configuration).
1433 * @param string $url Optional URL, to extract the domain from for rel =>
1434 * nofollow if appropriate
1435 * @return array Associative array of HTML attributes
1437 function getExternalLinkAttribs( $url = false ) {
1439 global $wgNoFollowLinks, $wgNoFollowNsExceptions;
1440 $ns = $this->mTitle
->getNamespace();
1441 if( $wgNoFollowLinks && !in_array($ns, $wgNoFollowNsExceptions) ) {
1442 $attribs['rel'] = 'nofollow';
1444 global $wgNoFollowDomainExceptions;
1445 if ( $wgNoFollowDomainExceptions ) {
1446 $bits = wfParseUrl( $url );
1447 if ( is_array( $bits ) && isset( $bits['host'] ) ) {
1448 foreach ( $wgNoFollowDomainExceptions as $domain ) {
1449 if( substr( $bits['host'], -strlen( $domain ) )
1451 unset( $attribs['rel'] );
1458 if ( $this->mOptions
->getExternalLinkTarget() ) {
1459 $attribs['target'] = $this->mOptions
->getExternalLinkTarget();
1466 * Replace unusual URL escape codes with their equivalent characters
1470 * @todo This can merge genuinely required bits in the path or query string,
1471 * breaking legit URLs. A proper fix would treat the various parts of
1472 * the URL differently; as a workaround, just use the output for
1473 * statistical records, not for actual linking/output.
1475 static function replaceUnusualEscapes( $url ) {
1476 return preg_replace_callback( '/%[0-9A-Fa-f]{2}/',
1477 array( __CLASS__
, 'replaceUnusualEscapesCallback' ), $url );
1481 * Callback function used in replaceUnusualEscapes().
1482 * Replaces unusual URL escape codes with their equivalent character
1486 private static function replaceUnusualEscapesCallback( $matches ) {
1487 $char = urldecode( $matches[0] );
1488 $ord = ord( $char );
1489 // Is it an unsafe or HTTP reserved character according to RFC 1738?
1490 if ( $ord > 32 && $ord < 127 && strpos( '<>"#{}|\^~[]`;/?', $char ) === false ) {
1491 // No, shouldn't be escaped
1494 // Yes, leave it escaped
1500 * make an image if it's allowed, either through the global
1501 * option, through the exception, or through the on-wiki whitelist
1504 function maybeMakeExternalImage( $url ) {
1505 $sk = $this->mOptions
->getSkin();
1506 $imagesfrom = $this->mOptions
->getAllowExternalImagesFrom();
1507 $imagesexception = !empty($imagesfrom);
1509 # $imagesfrom could be either a single string or an array of strings, parse out the latter
1510 if( $imagesexception && is_array( $imagesfrom ) ) {
1511 $imagematch = false;
1512 foreach( $imagesfrom as $match ) {
1513 if( strpos( $url, $match ) === 0 ) {
1518 } elseif( $imagesexception ) {
1519 $imagematch = (strpos( $url, $imagesfrom ) === 0);
1521 $imagematch = false;
1523 if ( $this->mOptions
->getAllowExternalImages()
1524 ||
( $imagesexception && $imagematch ) ) {
1525 if ( preg_match( self
::EXT_IMAGE_REGEX
, $url ) ) {
1527 $text = $sk->makeExternalImage( $url );
1530 if( !$text && $this->mOptions
->getEnableImageWhitelist()
1531 && preg_match( self
::EXT_IMAGE_REGEX
, $url ) ) {
1532 $whitelist = explode( "\n", wfMsgForContent( 'external_image_whitelist' ) );
1533 foreach( $whitelist as $entry ) {
1534 # Sanitize the regex fragment, make it case-insensitive, ignore blank entries/comments
1535 if( strpos( $entry, '#' ) === 0 ||
$entry === '' )
1537 if( preg_match( '/' . str_replace( '/', '\\/', $entry ) . '/i', $url ) ) {
1538 # Image matches a whitelist entry
1539 $text = $sk->makeExternalImage( $url );
1548 * Process [[ ]] wikilinks
1549 * @return processed text
1553 function replaceInternalLinks( $s ) {
1554 $this->mLinkHolders
->merge( $this->replaceInternalLinks2( $s ) );
1559 * Process [[ ]] wikilinks (RIL)
1560 * @return LinkHolderArray
1564 function replaceInternalLinks2( &$s ) {
1567 wfProfileIn( __METHOD__
);
1569 wfProfileIn( __METHOD__
.'-setup' );
1570 static $tc = FALSE, $e1, $e1_img;
1571 # the % is needed to support urlencoded titles as well
1573 $tc = Title
::legalChars() . '#%';
1574 # Match a link having the form [[namespace:link|alternate]]trail
1575 $e1 = "/^([{$tc}]+)(?:\\|(.+?))?]](.*)\$/sD";
1576 # Match cases where there is no "]]", which might still be images
1577 $e1_img = "/^([{$tc}]+)\\|(.*)\$/sD";
1580 $sk = $this->mOptions
->getSkin();
1581 $holders = new LinkHolderArray( $this );
1583 #split the entire text string on occurences of [[
1584 $a = StringUtils
::explode( '[[', ' ' . $s );
1585 #get the first element (all text up to first [[), and remove the space we added
1588 $line = $a->current(); # Workaround for broken ArrayIterator::next() that returns "void"
1589 $s = substr( $s, 1 );
1591 $useLinkPrefixExtension = $wgContLang->linkPrefixExtension();
1593 if ( $useLinkPrefixExtension ) {
1594 # Match the end of a line for a word that's not followed by whitespace,
1595 # e.g. in the case of 'The Arab al[[Razi]]', 'al' will be matched
1596 $e2 = wfMsgForContent( 'linkprefix' );
1599 if( is_null( $this->mTitle
) ) {
1600 wfProfileOut( __METHOD__
.'-setup' );
1601 wfProfileOut( __METHOD__
);
1602 throw new MWException( __METHOD__
.": \$this->mTitle is null\n" );
1604 $nottalk = !$this->mTitle
->isTalkPage();
1606 if ( $useLinkPrefixExtension ) {
1608 if ( preg_match( $e2, $s, $m ) ) {
1609 $first_prefix = $m[2];
1611 $first_prefix = false;
1617 if($wgContLang->hasVariants()) {
1618 $selflink = $wgContLang->convertLinkToAllVariants($this->mTitle
->getPrefixedText());
1620 $selflink = array($this->mTitle
->getPrefixedText());
1622 $useSubpages = $this->areSubpagesAllowed();
1623 wfProfileOut( __METHOD__
.'-setup' );
1625 # Loop for each link
1626 for ( ; $line !== false && $line !== null ; $a->next(), $line = $a->current() ) {
1627 # Check for excessive memory usage
1628 if ( $holders->isBig() ) {
1630 # Do the existence check, replace the link holders and clear the array
1631 $holders->replace( $s );
1635 if ( $useLinkPrefixExtension ) {
1636 wfProfileIn( __METHOD__
.'-prefixhandling' );
1637 if ( preg_match( $e2, $s, $m ) ) {
1645 $prefix = $first_prefix;
1646 $first_prefix = false;
1648 wfProfileOut( __METHOD__
.'-prefixhandling' );
1651 $might_be_img = false;
1653 wfProfileIn( __METHOD__
."-e1" );
1654 if ( preg_match( $e1, $line, $m ) ) { # page with normal text or alt
1656 # If we get a ] at the beginning of $m[3] that means we have a link that's something like:
1657 # [[Image:Foo.jpg|[http://example.com desc]]] <- having three ] in a row fucks up,
1658 # the real problem is with the $e1 regex
1661 # Still some problems for cases where the ] is meant to be outside punctuation,
1662 # and no image is in sight. See bug 2095.
1665 substr( $m[3], 0, 1 ) === ']' &&
1666 strpos($text, '[') !== false
1669 $text .= ']'; # so that replaceExternalLinks($text) works later
1670 $m[3] = substr( $m[3], 1 );
1672 # fix up urlencoded title texts
1673 if( strpos( $m[1], '%' ) !== false ) {
1674 # Should anchors '#' also be rejected?
1675 $m[1] = str_replace( array('<', '>'), array('<', '>'), urldecode($m[1]) );
1678 } elseif( preg_match($e1_img, $line, $m) ) { # Invalid, but might be an image with a link in its caption
1679 $might_be_img = true;
1681 if ( strpos( $m[1], '%' ) !== false ) {
1682 $m[1] = urldecode($m[1]);
1685 } else { # Invalid form; output directly
1686 $s .= $prefix . '[[' . $line ;
1687 wfProfileOut( __METHOD__
."-e1" );
1690 wfProfileOut( __METHOD__
."-e1" );
1691 wfProfileIn( __METHOD__
."-misc" );
1693 # Don't allow internal links to pages containing
1694 # PROTO: where PROTO is a valid URL protocol; these
1695 # should be external links.
1696 if (preg_match('/^\b(?:' . wfUrlProtocols() . ')/', $m[1])) {
1697 $s .= $prefix . '[[' . $line ;
1698 wfProfileOut( __METHOD__
."-misc" );
1702 # Make subpage if necessary
1703 if( $useSubpages ) {
1704 $link = $this->maybeDoSubpageLink( $m[1], $text );
1709 $noforce = (substr($m[1], 0, 1) !== ':');
1711 # Strip off leading ':'
1712 $link = substr($link, 1);
1715 wfProfileOut( __METHOD__
."-misc" );
1716 wfProfileIn( __METHOD__
."-title" );
1717 $nt = Title
::newFromText( $this->mStripState
->unstripNoWiki($link) );
1719 $s .= $prefix . '[[' . $line;
1720 wfProfileOut( __METHOD__
."-title" );
1724 $ns = $nt->getNamespace();
1725 $iw = $nt->getInterWiki();
1726 wfProfileOut( __METHOD__
."-title" );
1728 if ($might_be_img) { # if this is actually an invalid link
1729 wfProfileIn( __METHOD__
."-might_be_img" );
1730 if ($ns == NS_FILE
&& $noforce) { #but might be an image
1733 #look at the next 'line' to see if we can close it there
1735 $next_line = $a->current();
1736 if ( $next_line === false ||
$next_line === null ) {
1739 $m = explode( ']]', $next_line, 3 );
1740 if ( count( $m ) == 3 ) {
1741 # the first ]] closes the inner link, the second the image
1743 $text .= "[[{$m[0]}]]{$m[1]}";
1746 } elseif ( count( $m ) == 2 ) {
1747 #if there's exactly one ]] that's fine, we'll keep looking
1748 $text .= "[[{$m[0]}]]{$m[1]}";
1750 #if $next_line is invalid too, we need look no further
1751 $text .= '[[' . $next_line;
1756 # we couldn't find the end of this imageLink, so output it raw
1757 #but don't ignore what might be perfectly normal links in the text we've examined
1758 $holders->merge( $this->replaceInternalLinks2( $text ) );
1759 $s .= "{$prefix}[[$link|$text";
1760 # note: no $trail, because without an end, there *is* no trail
1761 wfProfileOut( __METHOD__
."-might_be_img" );
1764 } else { #it's not an image, so output it raw
1765 $s .= "{$prefix}[[$link|$text";
1766 # note: no $trail, because without an end, there *is* no trail
1767 wfProfileOut( __METHOD__
."-might_be_img" );
1770 wfProfileOut( __METHOD__
."-might_be_img" );
1773 $wasblank = ( '' == $text );
1774 if( $wasblank ) $text = $link;
1776 # Link not escaped by : , create the various objects
1780 wfProfileIn( __METHOD__
."-interwiki" );
1781 if( $iw && $this->mOptions
->getInterwikiMagic() && $nottalk && $wgContLang->getLanguageName( $iw ) ) {
1782 $this->mOutput
->addLanguageLink( $nt->getFullText() );
1783 $s = rtrim($s . $prefix);
1784 $s .= trim($trail, "\n") == '' ?
'': $prefix . $trail;
1785 wfProfileOut( __METHOD__
."-interwiki" );
1788 wfProfileOut( __METHOD__
."-interwiki" );
1790 if ( $ns == NS_FILE
) {
1791 wfProfileIn( __METHOD__
."-image" );
1792 if ( !wfIsBadImage( $nt->getDBkey(), $this->mTitle
) ) {
1793 # recursively parse links inside the image caption
1794 # actually, this will parse them in any other parameters, too,
1795 # but it might be hard to fix that, and it doesn't matter ATM
1796 $text = $this->replaceExternalLinks($text);
1797 $holders->merge( $this->replaceInternalLinks2( $text ) );
1799 # cloak any absolute URLs inside the image markup, so replaceExternalLinks() won't touch them
1800 $s .= $prefix . $this->armorLinks( $this->makeImage( $nt, $text, $holders ) ) . $trail;
1802 $this->mOutput
->addImage( $nt->getDBkey() );
1803 wfProfileOut( __METHOD__
."-image" );
1808 if ( $ns == NS_CATEGORY
) {
1809 wfProfileIn( __METHOD__
."-category" );
1810 $s = rtrim($s . "\n"); # bug 87
1813 $sortkey = $this->getDefaultSort();
1817 $sortkey = Sanitizer
::decodeCharReferences( $sortkey );
1818 $sortkey = str_replace( "\n", '', $sortkey );
1819 $sortkey = $wgContLang->convertCategoryKey( $sortkey );
1820 $this->mOutput
->addCategory( $nt->getDBkey(), $sortkey );
1823 * Strip the whitespace Category links produce, see bug 87
1824 * @todo We might want to use trim($tmp, "\n") here.
1826 $s .= trim($prefix . $trail, "\n") == '' ?
'': $prefix . $trail;
1828 wfProfileOut( __METHOD__
."-category" );
1833 # Self-link checking
1834 if( $nt->getFragment() === '' && $ns != NS_SPECIAL
) {
1835 if( in_array( $nt->getPrefixedText(), $selflink, true ) ) {
1836 $s .= $prefix . $sk->makeSelfLinkObj( $nt, $text, '', $trail );
1841 # NS_MEDIA is a pseudo-namespace for linking directly to a file
1842 # FIXME: Should do batch file existence checks, see comment below
1843 if( $ns == NS_MEDIA
) {
1844 # Give extensions a chance to select the file revision for us
1845 $skip = $time = false;
1846 wfRunHooks( 'BeforeParserMakeImageLinkObj', array( &$this, &$nt, &$skip, &$time ) );
1848 $link = $sk->link( $nt );
1850 $link = $sk->makeMediaLinkObj( $nt, $text, $time );
1852 # Cloak with NOPARSE to avoid replacement in replaceExternalLinks
1853 $s .= $prefix . $this->armorLinks( $link ) . $trail;
1854 $this->mOutput
->addImage( $nt->getDBkey() );
1858 # Some titles, such as valid special pages or files in foreign repos, should
1859 # be shown as bluelinks even though they're not included in the page table
1861 # FIXME: isAlwaysKnown() can be expensive for file links; we should really do
1862 # batch file existence checks for NS_FILE and NS_MEDIA
1863 if( $iw == '' && $nt->isAlwaysKnown() ) {
1864 $this->mOutput
->addLink( $nt );
1865 $s .= $this->makeKnownLinkHolder( $nt, $text, '', $trail, $prefix );
1867 # Links will be added to the output link list after checking
1868 $s .= $holders->makeHolder( $nt, $text, '', $trail, $prefix );
1871 wfProfileOut( __METHOD__
);
1876 * Make a link placeholder. The text returned can be later resolved to a real link with
1877 * replaceLinkHolders(). This is done for two reasons: firstly to avoid further
1878 * parsing of interwiki links, and secondly to allow all existence checks and
1879 * article length checks (for stub links) to be bundled into a single query.
1883 function makeLinkHolder( &$nt, $text = '', $query = '', $trail = '', $prefix = '' ) {
1884 return $this->mLinkHolders
->makeHolder( $nt, $text, $query, $trail, $prefix );
1888 * Render a forced-blue link inline; protect against double expansion of
1889 * URLs if we're in a mode that prepends full URL prefixes to internal links.
1890 * Since this little disaster has to split off the trail text to avoid
1891 * breaking URLs in the following text without breaking trails on the
1892 * wiki links, it's been made into a horrible function.
1895 * @param string $text
1896 * @param string $query
1897 * @param string $trail
1898 * @param string $prefix
1899 * @return string HTML-wikitext mix oh yuck
1901 function makeKnownLinkHolder( $nt, $text = '', $query = '', $trail = '', $prefix = '' ) {
1902 list( $inside, $trail ) = Linker
::splitTrail( $trail );
1903 $sk = $this->mOptions
->getSkin();
1904 $link = $sk->makeKnownLinkObj( $nt, $text, $query, $inside, $prefix );
1905 return $this->armorLinks( $link ) . $trail;
1909 * Insert a NOPARSE hacky thing into any inline links in a chunk that's
1910 * going to go through further parsing steps before inline URL expansion.
1912 * Not needed quite as much as it used to be since free links are a bit
1913 * more sensible these days. But bracketed links are still an issue.
1915 * @param string more-or-less HTML
1916 * @return string less-or-more HTML with NOPARSE bits
1918 function armorLinks( $text ) {
1919 return preg_replace( '/\b(' . wfUrlProtocols() . ')/',
1920 "{$this->mUniqPrefix}NOPARSE$1", $text );
1924 * Return true if subpage links should be expanded on this page.
1927 function areSubpagesAllowed() {
1928 # Some namespaces don't allow subpages
1929 return MWNamespace
::hasSubpages( $this->mTitle
->getNamespace() );
1933 * Handle link to subpage if necessary
1934 * @param string $target the source of the link
1935 * @param string &$text the link text, modified as necessary
1936 * @return string the full name of the link
1939 function maybeDoSubpageLink($target, &$text) {
1942 # :Foobar -- override special treatment of prefix (images, language links)
1943 # /Foobar -- convert to CurrentPage/Foobar
1944 # /Foobar/ -- convert to CurrentPage/Foobar, strip the initial / from text
1945 # ../ -- convert to CurrentPage, from CurrentPage/CurrentSubPage
1946 # ../Foobar -- convert to CurrentPage/Foobar, from CurrentPage/CurrentSubPage
1948 wfProfileIn( __METHOD__
);
1949 $ret = $target; # default return value is no change
1951 # Some namespaces don't allow subpages,
1952 # so only perform processing if subpages are allowed
1953 if( $this->areSubpagesAllowed() ) {
1954 $hash = strpos( $target, '#' );
1955 if( $hash !== false ) {
1956 $suffix = substr( $target, $hash );
1957 $target = substr( $target, 0, $hash );
1962 $target = trim( $target );
1963 # Look at the first character
1964 if( $target != '' && $target{0} === '/' ) {
1965 # / at end means we don't want the slash to be shown
1967 $trailingSlashes = preg_match_all( '%(/+)$%', $target, $m );
1968 if( $trailingSlashes ) {
1969 $noslash = $target = substr( $target, 1, -strlen($m[0][0]) );
1971 $noslash = substr( $target, 1 );
1974 $ret = $this->mTitle
->getPrefixedText(). '/' . trim($noslash) . $suffix;
1975 if( '' === $text ) {
1976 $text = $target . $suffix;
1977 } # this might be changed for ugliness reasons
1979 # check for .. subpage backlinks
1981 $nodotdot = $target;
1982 while( strncmp( $nodotdot, "../", 3 ) == 0 ) {
1984 $nodotdot = substr( $nodotdot, 3 );
1986 if($dotdotcount > 0) {
1987 $exploded = explode( '/', $this->mTitle
->GetPrefixedText() );
1988 if( count( $exploded ) > $dotdotcount ) { # not allowed to go below top level page
1989 $ret = implode( '/', array_slice( $exploded, 0, -$dotdotcount ) );
1990 # / at the end means don't show full path
1991 if( substr( $nodotdot, -1, 1 ) === '/' ) {
1992 $nodotdot = substr( $nodotdot, 0, -1 );
1993 if( '' === $text ) {
1994 $text = $nodotdot . $suffix;
1997 $nodotdot = trim( $nodotdot );
1998 if( $nodotdot != '' ) {
1999 $ret .= '/' . $nodotdot;
2007 wfProfileOut( __METHOD__
);
2012 * Used by doBlockLevels()
2015 /* private */ function closeParagraph() {
2017 if ( '' != $this->mLastSection
) {
2018 $result = '</' . $this->mLastSection
. ">\n";
2020 $this->mInPre
= false;
2021 $this->mLastSection
= '';
2024 # getCommon() returns the length of the longest common substring
2025 # of both arguments, starting at the beginning of both.
2027 /* private */ function getCommon( $st1, $st2 ) {
2028 $fl = strlen( $st1 );
2029 $shorter = strlen( $st2 );
2030 if ( $fl < $shorter ) { $shorter = $fl; }
2032 for ( $i = 0; $i < $shorter; ++
$i ) {
2033 if ( $st1{$i} != $st2{$i} ) { break; }
2037 # These next three functions open, continue, and close the list
2038 # element appropriate to the prefix character passed into them.
2040 /* private */ function openList( $char ) {
2041 $result = $this->closeParagraph();
2043 if ( '*' === $char ) { $result .= '<ul><li>'; }
2044 else if ( '#' === $char ) { $result .= '<ol><li>'; }
2045 else if ( ':' === $char ) { $result .= '<dl><dd>'; }
2046 else if ( ';' === $char ) {
2047 $result .= '<dl><dt>';
2048 $this->mDTopen
= true;
2050 else { $result = '<!-- ERR 1 -->'; }
2055 /* private */ function nextItem( $char ) {
2056 if ( '*' === $char ||
'#' === $char ) { return '</li><li>'; }
2057 else if ( ':' === $char ||
';' === $char ) {
2059 if ( $this->mDTopen
) { $close = '</dt>'; }
2060 if ( ';' === $char ) {
2061 $this->mDTopen
= true;
2062 return $close . '<dt>';
2064 $this->mDTopen
= false;
2065 return $close . '<dd>';
2068 return '<!-- ERR 2 -->';
2071 /* private */ function closeList( $char ) {
2072 if ( '*' === $char ) { $text = '</li></ul>'; }
2073 else if ( '#' === $char ) { $text = '</li></ol>'; }
2074 else if ( ':' === $char ) {
2075 if ( $this->mDTopen
) {
2076 $this->mDTopen
= false;
2077 $text = '</dt></dl>';
2079 $text = '</dd></dl>';
2082 else { return '<!-- ERR 3 -->'; }
2088 * Make lists from lines starting with ':', '*', '#', etc. (DBL)
2091 * @return string the lists rendered as HTML
2093 function doBlockLevels( $text, $linestart ) {
2094 wfProfileIn( __METHOD__
);
2096 # Parsing through the text line by line. The main thing
2097 # happening here is handling of block-level elements p, pre,
2098 # and making lists from lines starting with * # : etc.
2100 $textLines = StringUtils
::explode( "\n", $text );
2102 $lastPrefix = $output = '';
2103 $this->mDTopen
= $inBlockElem = false;
2105 $paragraphStack = false;
2107 foreach ( $textLines as $oLine ) {
2109 if ( !$linestart ) {
2115 $lastPrefixLength = strlen( $lastPrefix );
2116 $preCloseMatch = preg_match('/<\\/pre/i', $oLine );
2117 $preOpenMatch = preg_match('/<pre/i', $oLine );
2118 if ( !$this->mInPre
) {
2119 # Multiple prefixes may abut each other for nested lists.
2120 $prefixLength = strspn( $oLine, '*#:;' );
2121 $prefix = substr( $oLine, 0, $prefixLength );
2124 $prefix2 = str_replace( ';', ':', $prefix );
2125 $t = substr( $oLine, $prefixLength );
2126 $this->mInPre
= (bool)$preOpenMatch;
2128 # Don't interpret any other prefixes in preformatted text
2130 $prefix = $prefix2 = '';
2135 if( $prefixLength && $lastPrefix === $prefix2 ) {
2136 # Same as the last item, so no need to deal with nesting or opening stuff
2137 $output .= $this->nextItem( substr( $prefix, -1 ) );
2138 $paragraphStack = false;
2140 if ( substr( $prefix, -1 ) === ';') {
2141 # The one nasty exception: definition lists work like this:
2142 # ; title : definition text
2143 # So we check for : in the remainder text to split up the
2144 # title and definition, without b0rking links.
2146 if ($this->findColonNoLinks($t, $term, $t2) !== false) {
2148 $output .= $term . $this->nextItem( ':' );
2151 } elseif( $prefixLength ||
$lastPrefixLength ) {
2152 # Either open or close a level...
2153 $commonPrefixLength = $this->getCommon( $prefix, $lastPrefix );
2154 $paragraphStack = false;
2156 while( $commonPrefixLength < $lastPrefixLength ) {
2157 $output .= $this->closeList( $lastPrefix[$lastPrefixLength-1] );
2158 --$lastPrefixLength;
2160 if ( $prefixLength <= $commonPrefixLength && $commonPrefixLength > 0 ) {
2161 $output .= $this->nextItem( $prefix[$commonPrefixLength-1] );
2163 while ( $prefixLength > $commonPrefixLength ) {
2164 $char = substr( $prefix, $commonPrefixLength, 1 );
2165 $output .= $this->openList( $char );
2167 if ( ';' === $char ) {
2168 # FIXME: This is dupe of code above
2169 if ($this->findColonNoLinks($t, $term, $t2) !== false) {
2171 $output .= $term . $this->nextItem( ':' );
2174 ++
$commonPrefixLength;
2176 $lastPrefix = $prefix2;
2178 if( 0 == $prefixLength ) {
2179 wfProfileIn( __METHOD__
."-paragraph" );
2180 # No prefix (not in list)--go to paragraph mode
2181 // XXX: use a stack for nestable elements like span, table and div
2182 $openmatch = preg_match('/(?:<table|<blockquote|<h1|<h2|<h3|<h4|<h5|<h6|<pre|<tr|<p|<ul|<ol|<li|<\\/tr|<\\/td|<\\/th)/iS', $t );
2183 $closematch = preg_match(
2184 '/(?:<\\/table|<\\/blockquote|<\\/h1|<\\/h2|<\\/h3|<\\/h4|<\\/h5|<\\/h6|'.
2185 '<td|<th|<\\/?div|<hr|<\\/pre|<\\/p|'.$this->mUniqPrefix
.'-pre|<\\/li|<\\/ul|<\\/ol|<\\/?center)/iS', $t );
2186 if ( $openmatch or $closematch ) {
2187 $paragraphStack = false;
2188 #Â TODO bug 5718: paragraph closed
2189 $output .= $this->closeParagraph();
2190 if ( $preOpenMatch and !$preCloseMatch ) {
2191 $this->mInPre
= true;
2193 if ( $closematch ) {
2194 $inBlockElem = false;
2196 $inBlockElem = true;
2198 } else if ( !$inBlockElem && !$this->mInPre
) {
2199 if ( ' ' == $t{0} and ( $this->mLastSection
=== 'pre' or trim($t) != '' ) ) {
2201 if ($this->mLastSection
!== 'pre') {
2202 $paragraphStack = false;
2203 $output .= $this->closeParagraph().'<pre>';
2204 $this->mLastSection
= 'pre';
2206 $t = substr( $t, 1 );
2209 if ( '' == trim($t) ) {
2210 if ( $paragraphStack ) {
2211 $output .= $paragraphStack.'<br />';
2212 $paragraphStack = false;
2213 $this->mLastSection
= 'p';
2215 if ($this->mLastSection
!== 'p' ) {
2216 $output .= $this->closeParagraph();
2217 $this->mLastSection
= '';
2218 $paragraphStack = '<p>';
2220 $paragraphStack = '</p><p>';
2224 if ( $paragraphStack ) {
2225 $output .= $paragraphStack;
2226 $paragraphStack = false;
2227 $this->mLastSection
= 'p';
2228 } else if ($this->mLastSection
!== 'p') {
2229 $output .= $this->closeParagraph().'<p>';
2230 $this->mLastSection
= 'p';
2235 wfProfileOut( __METHOD__
."-paragraph" );
2237 // somewhere above we forget to get out of pre block (bug 785)
2238 if($preCloseMatch && $this->mInPre
) {
2239 $this->mInPre
= false;
2241 if ($paragraphStack === false) {
2245 while ( $prefixLength ) {
2246 $output .= $this->closeList( $prefix2[$prefixLength-1] );
2249 if ( '' != $this->mLastSection
) {
2250 $output .= '</' . $this->mLastSection
. '>';
2251 $this->mLastSection
= '';
2254 wfProfileOut( __METHOD__
);
2259 * Split up a string on ':', ignoring any occurences inside tags
2260 * to prevent illegal overlapping.
2261 * @param string $str the string to split
2262 * @param string &$before set to everything before the ':'
2263 * @param string &$after set to everything after the ':'
2264 * return string the position of the ':', or false if none found
2266 function findColonNoLinks($str, &$before, &$after) {
2267 wfProfileIn( __METHOD__
);
2269 $pos = strpos( $str, ':' );
2270 if( $pos === false ) {
2272 wfProfileOut( __METHOD__
);
2276 $lt = strpos( $str, '<' );
2277 if( $lt === false ||
$lt > $pos ) {
2278 // Easy; no tag nesting to worry about
2279 $before = substr( $str, 0, $pos );
2280 $after = substr( $str, $pos+
1 );
2281 wfProfileOut( __METHOD__
);
2285 // Ugly state machine to walk through avoiding tags.
2286 $state = self
::COLON_STATE_TEXT
;
2288 $len = strlen( $str );
2289 for( $i = 0; $i < $len; $i++
) {
2293 // (Using the number is a performance hack for common cases)
2294 case 0: // self::COLON_STATE_TEXT:
2297 // Could be either a <start> tag or an </end> tag
2298 $state = self
::COLON_STATE_TAGSTART
;
2303 $before = substr( $str, 0, $i );
2304 $after = substr( $str, $i +
1 );
2305 wfProfileOut( __METHOD__
);
2308 // Embedded in a tag; don't break it.
2311 // Skip ahead looking for something interesting
2312 $colon = strpos( $str, ':', $i );
2313 if( $colon === false ) {
2314 // Nothing else interesting
2315 wfProfileOut( __METHOD__
);
2318 $lt = strpos( $str, '<', $i );
2319 if( $stack === 0 ) {
2320 if( $lt === false ||
$colon < $lt ) {
2322 $before = substr( $str, 0, $colon );
2323 $after = substr( $str, $colon +
1 );
2324 wfProfileOut( __METHOD__
);
2328 if( $lt === false ) {
2329 // Nothing else interesting to find; abort!
2330 // We're nested, but there's no close tags left. Abort!
2333 // Skip ahead to next tag start
2335 $state = self
::COLON_STATE_TAGSTART
;
2338 case 1: // self::COLON_STATE_TAG:
2343 $state = self
::COLON_STATE_TEXT
;
2346 // Slash may be followed by >?
2347 $state = self
::COLON_STATE_TAGSLASH
;
2353 case 2: // self::COLON_STATE_TAGSTART:
2356 $state = self
::COLON_STATE_CLOSETAG
;
2359 $state = self
::COLON_STATE_COMMENT
;
2362 // Illegal early close? This shouldn't happen D:
2363 $state = self
::COLON_STATE_TEXT
;
2366 $state = self
::COLON_STATE_TAG
;
2369 case 3: // self::COLON_STATE_CLOSETAG:
2374 wfDebug( __METHOD__
.": Invalid input; too many close tags\n" );
2375 wfProfileOut( __METHOD__
);
2378 $state = self
::COLON_STATE_TEXT
;
2381 case self
::COLON_STATE_TAGSLASH
:
2383 // Yes, a self-closed tag <blah/>
2384 $state = self
::COLON_STATE_TEXT
;
2386 // Probably we're jumping the gun, and this is an attribute
2387 $state = self
::COLON_STATE_TAG
;
2390 case 5: // self::COLON_STATE_COMMENT:
2392 $state = self
::COLON_STATE_COMMENTDASH
;
2395 case self
::COLON_STATE_COMMENTDASH
:
2397 $state = self
::COLON_STATE_COMMENTDASHDASH
;
2399 $state = self
::COLON_STATE_COMMENT
;
2402 case self
::COLON_STATE_COMMENTDASHDASH
:
2404 $state = self
::COLON_STATE_TEXT
;
2406 $state = self
::COLON_STATE_COMMENT
;
2410 throw new MWException( "State machine error in " . __METHOD__
);
2414 wfDebug( __METHOD__
.": Invalid input; not enough close tags (stack $stack, state $state)\n" );
2417 wfProfileOut( __METHOD__
);
2422 * Return value of a magic variable (like PAGENAME)
2426 function getVariableValue( $index ) {
2427 global $wgContLang, $wgSitename, $wgServer, $wgServerName, $wgScriptPath;
2430 * Some of these require message or data lookups and can be
2431 * expensive to check many times.
2433 if ( wfRunHooks( 'ParserGetVariableValueVarCache', array( &$this, &$this->mVarCache
) ) ) {
2434 if ( isset( $this->mVarCache
[$index] ) ) {
2435 return $this->mVarCache
[$index];
2439 $ts = wfTimestamp( TS_UNIX
, $this->mOptions
->getTimestamp() );
2440 wfRunHooks( 'ParserGetVariableValueTs', array( &$this, &$ts ) );
2443 global $wgLocaltimezone;
2444 if ( isset( $wgLocaltimezone ) ) {
2445 $oldtz = getenv( 'TZ' );
2446 putenv( 'TZ='.$wgLocaltimezone );
2449 wfSuppressWarnings(); // E_STRICT system time bitching
2450 $localTimestamp = date( 'YmdHis', $ts );
2451 $localMonth = date( 'm', $ts );
2452 $localMonthName = date( 'n', $ts );
2453 $localDay = date( 'j', $ts );
2454 $localDay2 = date( 'd', $ts );
2455 $localDayOfWeek = date( 'w', $ts );
2456 $localWeek = date( 'W', $ts );
2457 $localYear = date( 'Y', $ts );
2458 $localHour = date( 'H', $ts );
2459 if ( isset( $wgLocaltimezone ) ) {
2460 putenv( 'TZ='.$oldtz );
2462 wfRestoreWarnings();
2465 case 'currentmonth':
2466 return $this->mVarCache
[$index] = $wgContLang->formatNum( gmdate( 'm', $ts ) );
2467 case 'currentmonthname':
2468 return $this->mVarCache
[$index] = $wgContLang->getMonthName( gmdate( 'n', $ts ) );
2469 case 'currentmonthnamegen':
2470 return $this->mVarCache
[$index] = $wgContLang->getMonthNameGen( gmdate( 'n', $ts ) );
2471 case 'currentmonthabbrev':
2472 return $this->mVarCache
[$index] = $wgContLang->getMonthAbbreviation( gmdate( 'n', $ts ) );
2474 return $this->mVarCache
[$index] = $wgContLang->formatNum( gmdate( 'j', $ts ) );
2476 return $this->mVarCache
[$index] = $wgContLang->formatNum( gmdate( 'd', $ts ) );
2478 return $this->mVarCache
[$index] = $wgContLang->formatNum( $localMonth );
2479 case 'localmonthname':
2480 return $this->mVarCache
[$index] = $wgContLang->getMonthName( $localMonthName );
2481 case 'localmonthnamegen':
2482 return $this->mVarCache
[$index] = $wgContLang->getMonthNameGen( $localMonthName );
2483 case 'localmonthabbrev':
2484 return $this->mVarCache
[$index] = $wgContLang->getMonthAbbreviation( $localMonthName );
2486 return $this->mVarCache
[$index] = $wgContLang->formatNum( $localDay );
2488 return $this->mVarCache
[$index] = $wgContLang->formatNum( $localDay2 );
2490 return wfEscapeWikiText( $this->mTitle
->getText() );
2492 return $this->mTitle
->getPartialURL();
2493 case 'fullpagename':
2494 return wfEscapeWikiText( $this->mTitle
->getPrefixedText() );
2495 case 'fullpagenamee':
2496 return $this->mTitle
->getPrefixedURL();
2498 return wfEscapeWikiText( $this->mTitle
->getSubpageText() );
2499 case 'subpagenamee':
2500 return $this->mTitle
->getSubpageUrlForm();
2501 case 'basepagename':
2502 return wfEscapeWikiText( $this->mTitle
->getBaseText() );
2503 case 'basepagenamee':
2504 return wfUrlEncode( str_replace( ' ', '_', $this->mTitle
->getBaseText() ) );
2505 case 'talkpagename':
2506 if( $this->mTitle
->canTalk() ) {
2507 $talkPage = $this->mTitle
->getTalkPage();
2508 return wfEscapeWikiText( $talkPage->getPrefixedText() );
2512 case 'talkpagenamee':
2513 if( $this->mTitle
->canTalk() ) {
2514 $talkPage = $this->mTitle
->getTalkPage();
2515 return $talkPage->getPrefixedUrl();
2519 case 'subjectpagename':
2520 $subjPage = $this->mTitle
->getSubjectPage();
2521 return wfEscapeWikiText( $subjPage->getPrefixedText() );
2522 case 'subjectpagenamee':
2523 $subjPage = $this->mTitle
->getSubjectPage();
2524 return $subjPage->getPrefixedUrl();
2526 // Let the edit saving system know we should parse the page
2527 // *after* a revision ID has been assigned.
2528 $this->mOutput
->setFlag( 'vary-revision' );
2529 wfDebug( __METHOD__
. ": {{REVISIONID}} used, setting vary-revision...\n" );
2530 return $this->mRevisionId
;
2532 // Let the edit saving system know we should parse the page
2533 // *after* a revision ID has been assigned. This is for null edits.
2534 $this->mOutput
->setFlag( 'vary-revision' );
2535 wfDebug( __METHOD__
. ": {{REVISIONDAY}} used, setting vary-revision...\n" );
2536 return intval( substr( $this->getRevisionTimestamp(), 6, 2 ) );
2537 case 'revisionday2':
2538 // Let the edit saving system know we should parse the page
2539 // *after* a revision ID has been assigned. This is for null edits.
2540 $this->mOutput
->setFlag( 'vary-revision' );
2541 wfDebug( __METHOD__
. ": {{REVISIONDAY2}} used, setting vary-revision...\n" );
2542 return substr( $this->getRevisionTimestamp(), 6, 2 );
2543 case 'revisionmonth':
2544 // Let the edit saving system know we should parse the page
2545 // *after* a revision ID has been assigned. This is for null edits.
2546 $this->mOutput
->setFlag( 'vary-revision' );
2547 wfDebug( __METHOD__
. ": {{REVISIONMONTH}} used, setting vary-revision...\n" );
2548 return intval( substr( $this->getRevisionTimestamp(), 4, 2 ) );
2549 case 'revisionyear':
2550 // Let the edit saving system know we should parse the page
2551 // *after* a revision ID has been assigned. This is for null edits.
2552 $this->mOutput
->setFlag( 'vary-revision' );
2553 wfDebug( __METHOD__
. ": {{REVISIONYEAR}} used, setting vary-revision...\n" );
2554 return substr( $this->getRevisionTimestamp(), 0, 4 );
2555 case 'revisiontimestamp':
2556 // Let the edit saving system know we should parse the page
2557 // *after* a revision ID has been assigned. This is for null edits.
2558 $this->mOutput
->setFlag( 'vary-revision' );
2559 wfDebug( __METHOD__
. ": {{REVISIONTIMESTAMP}} used, setting vary-revision...\n" );
2560 return $this->getRevisionTimestamp();
2562 return str_replace('_',' ',$wgContLang->getNsText( $this->mTitle
->getNamespace() ) );
2564 return wfUrlencode( $wgContLang->getNsText( $this->mTitle
->getNamespace() ) );
2566 return $this->mTitle
->canTalk() ?
str_replace('_',' ',$this->mTitle
->getTalkNsText()) : '';
2568 return $this->mTitle
->canTalk() ?
wfUrlencode( $this->mTitle
->getTalkNsText() ) : '';
2569 case 'subjectspace':
2570 return $this->mTitle
->getSubjectNsText();
2571 case 'subjectspacee':
2572 return( wfUrlencode( $this->mTitle
->getSubjectNsText() ) );
2573 case 'currentdayname':
2574 return $this->mVarCache
[$index] = $wgContLang->getWeekdayName( gmdate( 'w', $ts ) +
1 );
2576 return $this->mVarCache
[$index] = $wgContLang->formatNum( gmdate( 'Y', $ts ), true );
2578 return $this->mVarCache
[$index] = $wgContLang->time( wfTimestamp( TS_MW
, $ts ), false, false );
2580 return $this->mVarCache
[$index] = $wgContLang->formatNum( gmdate( 'H', $ts ), true );
2582 // @bug 4594 PHP5 has it zero padded, PHP4 does not, cast to
2583 // int to remove the padding
2584 return $this->mVarCache
[$index] = $wgContLang->formatNum( (int)gmdate( 'W', $ts ) );
2586 return $this->mVarCache
[$index] = $wgContLang->formatNum( gmdate( 'w', $ts ) );
2587 case 'localdayname':
2588 return $this->mVarCache
[$index] = $wgContLang->getWeekdayName( $localDayOfWeek +
1 );
2590 return $this->mVarCache
[$index] = $wgContLang->formatNum( $localYear, true );
2592 return $this->mVarCache
[$index] = $wgContLang->time( $localTimestamp, false, false );
2594 return $this->mVarCache
[$index] = $wgContLang->formatNum( $localHour, true );
2596 // @bug 4594 PHP5 has it zero padded, PHP4 does not, cast to
2597 // int to remove the padding
2598 return $this->mVarCache
[$index] = $wgContLang->formatNum( (int)$localWeek );
2600 return $this->mVarCache
[$index] = $wgContLang->formatNum( $localDayOfWeek );
2601 case 'numberofarticles':
2602 return $this->mVarCache
[$index] = $wgContLang->formatNum( SiteStats
::articles() );
2603 case 'numberoffiles':
2604 return $this->mVarCache
[$index] = $wgContLang->formatNum( SiteStats
::images() );
2605 case 'numberofusers':
2606 return $this->mVarCache
[$index] = $wgContLang->formatNum( SiteStats
::users() );
2607 case 'numberofpages':
2608 return $this->mVarCache
[$index] = $wgContLang->formatNum( SiteStats
::pages() );
2609 case 'numberofadmins':
2610 return $this->mVarCache
[$index] = $wgContLang->formatNum( SiteStats
::numberingroup('sysop') );
2611 case 'numberofedits':
2612 return $this->mVarCache
[$index] = $wgContLang->formatNum( SiteStats
::edits() );
2613 case 'numberofviews':
2614 return $this->mVarCache
[$index] = $wgContLang->formatNum( SiteStats
::views() );
2615 case 'currenttimestamp':
2616 return $this->mVarCache
[$index] = wfTimestamp( TS_MW
, $ts );
2617 case 'localtimestamp':
2618 return $this->mVarCache
[$index] = $localTimestamp;
2619 case 'currentversion':
2620 return $this->mVarCache
[$index] = SpecialVersion
::getVersion();
2626 return $wgServerName;
2628 return $wgScriptPath;
2629 case 'directionmark':
2630 return $wgContLang->getDirMark();
2631 case 'contentlanguage':
2632 global $wgContLanguageCode;
2633 return $wgContLanguageCode;
2636 if ( wfRunHooks( 'ParserGetVariableValueSwitch', array( &$this, &$this->mVarCache
, &$index, &$ret ) ) )
2644 * initialise the magic variables (like CURRENTMONTHNAME)
2648 function initialiseVariables() {
2649 wfProfileIn( __METHOD__
);
2650 $variableIDs = MagicWord
::getVariableIDs();
2652 $this->mVariables
= new MagicWordArray( $variableIDs );
2653 wfProfileOut( __METHOD__
);
2657 * Preprocess some wikitext and return the document tree.
2658 * This is the ghost of replace_variables().
2660 * @param string $text The text to parse
2661 * @param integer flags Bitwise combination of:
2662 * self::PTD_FOR_INCLUSION Handle <noinclude>/<includeonly> as if the text is being
2663 * included. Default is to assume a direct page view.
2665 * The generated DOM tree must depend only on the input text and the flags.
2666 * The DOM tree must be the same in OT_HTML and OT_WIKI mode, to avoid a regression of bug 4899.
2668 * Any flag added to the $flags parameter here, or any other parameter liable to cause a
2669 * change in the DOM tree for a given text, must be passed through the section identifier
2670 * in the section edit link and thus back to extractSections().
2672 * The output of this function is currently only cached in process memory, but a persistent
2673 * cache may be implemented at a later date which takes further advantage of these strict
2674 * dependency requirements.
2678 function preprocessToDom ( $text, $flags = 0 ) {
2679 $dom = $this->getPreprocessor()->preprocessToObj( $text, $flags );
2684 * Return a three-element array: leading whitespace, string contents, trailing whitespace
2686 public static function splitWhitespace( $s ) {
2687 $ltrimmed = ltrim( $s );
2688 $w1 = substr( $s, 0, strlen( $s ) - strlen( $ltrimmed ) );
2689 $trimmed = rtrim( $ltrimmed );
2690 $diff = strlen( $ltrimmed ) - strlen( $trimmed );
2692 $w2 = substr( $ltrimmed, -$diff );
2696 return array( $w1, $trimmed, $w2 );
2700 * Replace magic variables, templates, and template arguments
2701 * with the appropriate text. Templates are substituted recursively,
2702 * taking care to avoid infinite loops.
2704 * Note that the substitution depends on value of $mOutputType:
2705 * self::OT_WIKI: only {{subst:}} templates
2706 * self::OT_PREPROCESS: templates but not extension tags
2707 * self::OT_HTML: all templates and extension tags
2709 * @param string $tex The text to transform
2710 * @param PPFrame $frame Object describing the arguments passed to the template.
2711 * Arguments may also be provided as an associative array, as was the usual case before MW1.12.
2712 * Providing arguments this way may be useful for extensions wishing to perform variable replacement explicitly.
2713 * @param bool $argsOnly Only do argument (triple-brace) expansion, not double-brace expansion
2716 function replaceVariables( $text, $frame = false, $argsOnly = false ) {
2717 # Prevent too big inclusions
2718 if( strlen( $text ) > $this->mOptions
->getMaxIncludeSize() ) {
2722 wfProfileIn( __METHOD__
);
2724 if ( $frame === false ) {
2725 $frame = $this->getPreprocessor()->newFrame();
2726 } elseif ( !( $frame instanceof PPFrame
) ) {
2727 wfDebug( __METHOD__
." called using plain parameters instead of a PPFrame instance. Creating custom frame.\n" );
2728 $frame = $this->getPreprocessor()->newCustomFrame($frame);
2731 $dom = $this->preprocessToDom( $text );
2732 $flags = $argsOnly ? PPFrame
::NO_TEMPLATES
: 0;
2733 $text = $frame->expand( $dom, $flags );
2735 wfProfileOut( __METHOD__
);
2739 /// Clean up argument array - refactored in 1.9 so parserfunctions can use it, too.
2740 static function createAssocArgs( $args ) {
2741 $assocArgs = array();
2743 foreach( $args as $arg ) {
2744 $eqpos = strpos( $arg, '=' );
2745 if ( $eqpos === false ) {
2746 $assocArgs[$index++
] = $arg;
2748 $name = trim( substr( $arg, 0, $eqpos ) );
2749 $value = trim( substr( $arg, $eqpos+
1 ) );
2750 if ( $value === false ) {
2753 if ( $name !== false ) {
2754 $assocArgs[$name] = $value;
2763 * Warn the user when a parser limitation is reached
2764 * Will warn at most once the user per limitation type
2766 * @param string $limitationType, should be one of:
2767 * 'expensive-parserfunction' (corresponding messages: 'expensive-parserfunction-warning', 'expensive-parserfunction-category')
2768 * 'post-expand-template-argument' (corresponding messages: 'post-expand-template-argument-warning', 'post-expand-template-argument-category')
2769 * 'post-expand-template-inclusion' (corresponding messages: 'post-expand-template-inclusion-warning', 'post-expand-template-inclusion-category')
2770 * @params int $current, $max When an explicit limit has been
2771 * exceeded, provide the values (optional)
2773 function limitationWarn( $limitationType, $current=null, $max=null) {
2774 $msgName = $limitationType . '-warning';
2775 //does no harm if $current and $max are present but are unnecessary for the message
2776 $warning = wfMsgExt( $msgName, array( 'parsemag', 'escape' ), $current, $max );
2777 $this->mOutput
->addWarning( $warning );
2778 $cat = Title
::makeTitleSafe( NS_CATEGORY
, wfMsgForContent( $limitationType . '-category' ) );
2780 $this->mOutput
->addCategory( $cat->getDBkey(), $this->getDefaultSort() );
2785 * Return the text of a template, after recursively
2786 * replacing any variables or templates within the template.
2788 * @param array $piece The parts of the template
2789 * $piece['title']: the title, i.e. the part before the |
2790 * $piece['parts']: the parameter array
2791 * $piece['lineStart']: whether the brace was at the start of a line
2792 * @param PPFrame The current frame, contains template arguments
2793 * @return string the text of the template
2796 function braceSubstitution( $piece, $frame ) {
2797 global $wgContLang, $wgAllowDisplayTitle, $wgNonincludableNamespaces;
2798 wfProfileIn( __METHOD__
);
2799 wfProfileIn( __METHOD__
.'-setup' );
2802 $found = false; # $text has been filled
2803 $nowiki = false; # wiki markup in $text should be escaped
2804 $isHTML = false; # $text is HTML, armour it against wikitext transformation
2805 $forceRawInterwiki = false; # Force interwiki transclusion to be done in raw mode not rendered
2806 $isChildObj = false; # $text is a DOM node needing expansion in a child frame
2807 $isLocalObj = false; # $text is a DOM node needing expansion in the current frame
2809 # Title object, where $text came from
2812 # $part1 is the bit before the first |, and must contain only title characters.
2813 # Various prefixes will be stripped from it later.
2814 $titleWithSpaces = $frame->expand( $piece['title'] );
2815 $part1 = trim( $titleWithSpaces );
2818 # Original title text preserved for various purposes
2819 $originalTitle = $part1;
2821 # $args is a list of argument nodes, starting from index 0, not including $part1
2822 $args = (null == $piece['parts']) ?
array() : $piece['parts'];
2823 wfProfileOut( __METHOD__
.'-setup' );
2826 wfProfileIn( __METHOD__
.'-modifiers' );
2828 $mwSubst = MagicWord
::get( 'subst' );
2829 if ( $mwSubst->matchStartAndRemove( $part1 ) xor $this->ot
['wiki'] ) {
2830 # One of two possibilities is true:
2831 # 1) Found SUBST but not in the PST phase
2832 # 2) Didn't find SUBST and in the PST phase
2833 # In either case, return without further processing
2834 $text = $frame->virtualBracketedImplode( '{{', '|', '}}', $titleWithSpaces, $args );
2841 if ( !$found && $args->getLength() == 0 ) {
2842 $id = $this->mVariables
->matchStartToEnd( $part1 );
2843 if ( $id !== false ) {
2844 $text = $this->getVariableValue( $id );
2845 if (MagicWord
::getCacheTTL($id)>-1)
2846 $this->mOutput
->mContainsOldMagic
= true;
2851 # MSG, MSGNW and RAW
2854 $mwMsgnw = MagicWord
::get( 'msgnw' );
2855 if ( $mwMsgnw->matchStartAndRemove( $part1 ) ) {
2858 # Remove obsolete MSG:
2859 $mwMsg = MagicWord
::get( 'msg' );
2860 $mwMsg->matchStartAndRemove( $part1 );
2864 $mwRaw = MagicWord
::get( 'raw' );
2865 if ( $mwRaw->matchStartAndRemove( $part1 ) ) {
2866 $forceRawInterwiki = true;
2869 wfProfileOut( __METHOD__
.'-modifiers' );
2873 wfProfileIn( __METHOD__
. '-pfunc' );
2875 $colonPos = strpos( $part1, ':' );
2876 if ( $colonPos !== false ) {
2877 # Case sensitive functions
2878 $function = substr( $part1, 0, $colonPos );
2879 if ( isset( $this->mFunctionSynonyms
[1][$function] ) ) {
2880 $function = $this->mFunctionSynonyms
[1][$function];
2882 # Case insensitive functions
2883 $function = strtolower( $function );
2884 if ( isset( $this->mFunctionSynonyms
[0][$function] ) ) {
2885 $function = $this->mFunctionSynonyms
[0][$function];
2891 list( $callback, $flags ) = $this->mFunctionHooks
[$function];
2892 $initialArgs = array( &$this );
2893 $funcArgs = array( trim( substr( $part1, $colonPos +
1 ) ) );
2894 if ( $flags & SFH_OBJECT_ARGS
) {
2895 # Add a frame parameter, and pass the arguments as an array
2896 $allArgs = $initialArgs;
2897 $allArgs[] = $frame;
2898 for ( $i = 0; $i < $args->getLength(); $i++
) {
2899 $funcArgs[] = $args->item( $i );
2901 $allArgs[] = $funcArgs;
2903 # Convert arguments to plain text
2904 for ( $i = 0; $i < $args->getLength(); $i++
) {
2905 $funcArgs[] = trim( $frame->expand( $args->item( $i ) ) );
2907 $allArgs = array_merge( $initialArgs, $funcArgs );
2910 # Workaround for PHP bug 35229 and similar
2911 if ( !is_callable( $callback ) ) {
2912 throw new MWException( "Tag hook for $function is not callable\n" );
2914 $result = call_user_func_array( $callback, $allArgs );
2917 $preprocessFlags = 0;
2919 if ( is_array( $result ) ) {
2920 if ( isset( $result[0] ) ) {
2922 unset( $result[0] );
2925 // Extract flags into the local scope
2926 // This allows callers to set flags such as nowiki, found, etc.
2932 $text = $this->preprocessToDom( $text, $preprocessFlags );
2937 wfProfileOut( __METHOD__
. '-pfunc' );
2940 # Finish mangling title and then check for loops.
2941 # Set $title to a Title object and $titleText to the PDBK
2944 # Split the title into page and subpage
2946 $part1 = $this->maybeDoSubpageLink( $part1, $subpage );
2947 if ($subpage !== '') {
2948 $ns = $this->mTitle
->getNamespace();
2950 $title = Title
::newFromText( $part1, $ns );
2952 $titleText = $title->getPrefixedText();
2953 # Check for language variants if the template is not found
2954 if($wgContLang->hasVariants() && $title->getArticleID() == 0){
2955 $wgContLang->findVariantLink( $part1, $title, true );
2957 # Do infinite loop check
2958 if ( !$frame->loopCheck( $title ) ) {
2960 $text = '<span class="error">' . wfMsgForContent( 'parser-template-loop-warning', $titleText ) . '</span>';
2961 wfDebug( __METHOD__
.": template loop broken at '$titleText'\n" );
2963 # Do recursion depth check
2964 $limit = $this->mOptions
->getMaxTemplateDepth();
2965 if ( $frame->depth
>= $limit ) {
2967 $text = '<span class="error">' . wfMsgForContent( 'parser-template-recursion-depth-warning', $limit ) . '</span>';
2972 # Load from database
2973 if ( !$found && $title ) {
2974 wfProfileIn( __METHOD__
. '-loadtpl' );
2975 if ( !$title->isExternal() ) {
2976 if ( $title->getNamespace() == NS_SPECIAL
&& $this->mOptions
->getAllowSpecialInclusion() && $this->ot
['html'] ) {
2977 $text = SpecialPage
::capturePath( $title );
2978 if ( is_string( $text ) ) {
2981 $this->disableCache();
2983 } else if ( $wgNonincludableNamespaces && in_array( $title->getNamespace(), $wgNonincludableNamespaces ) ) {
2984 $found = false; //access denied
2985 wfDebug( __METHOD__
.": template inclusion denied for " . $title->getPrefixedDBkey() );
2987 list( $text, $title ) = $this->getTemplateDom( $title );
2988 if ( $text !== false ) {
2994 # If the title is valid but undisplayable, make a link to it
2995 if ( !$found && ( $this->ot
['html'] ||
$this->ot
['pre'] ) ) {
2996 $text = "[[:$titleText]]";
2999 } elseif ( $title->isTrans() ) {
3000 // Interwiki transclusion
3001 if ( $this->ot
['html'] && !$forceRawInterwiki ) {
3002 $text = $this->interwikiTransclude( $title, 'render' );
3005 $text = $this->interwikiTransclude( $title, 'raw' );
3006 // Preprocess it like a template
3007 $text = $this->preprocessToDom( $text, self
::PTD_FOR_INCLUSION
);
3012 wfProfileOut( __METHOD__
. '-loadtpl' );
3015 # If we haven't found text to substitute by now, we're done
3016 # Recover the source wikitext and return it
3018 $text = $frame->virtualBracketedImplode( '{{', '|', '}}', $titleWithSpaces, $args );
3019 wfProfileOut( __METHOD__
);
3020 return array( 'object' => $text );
3023 # Expand DOM-style return values in a child frame
3024 if ( $isChildObj ) {
3025 # Clean up argument array
3026 $newFrame = $frame->newChild( $args, $title );
3029 $text = $newFrame->expand( $text, PPFrame
::RECOVER_ORIG
);
3030 } elseif ( $titleText !== false && $newFrame->isEmpty() ) {
3031 # Expansion is eligible for the empty-frame cache
3032 if ( isset( $this->mTplExpandCache
[$titleText] ) ) {
3033 $text = $this->mTplExpandCache
[$titleText];
3035 $text = $newFrame->expand( $text );
3036 $this->mTplExpandCache
[$titleText] = $text;
3039 # Uncached expansion
3040 $text = $newFrame->expand( $text );
3043 if ( $isLocalObj && $nowiki ) {
3044 $text = $frame->expand( $text, PPFrame
::RECOVER_ORIG
);
3045 $isLocalObj = false;
3048 # Replace raw HTML by a placeholder
3049 # Add a blank line preceding, to prevent it from mucking up
3050 # immediately preceding headings
3052 $text = "\n\n" . $this->insertStripItem( $text );
3054 # Escape nowiki-style return values
3055 elseif ( $nowiki && ( $this->ot
['html'] ||
$this->ot
['pre'] ) ) {
3056 $text = wfEscapeWikiText( $text );
3058 # Bug 529: if the template begins with a table or block-level
3059 # element, it should be treated as beginning a new line.
3060 # This behaviour is somewhat controversial.
3061 elseif ( is_string( $text ) && !$piece['lineStart'] && preg_match('/^(?:{\\||:|;|#|\*)/', $text)) /*}*/{
3062 $text = "\n" . $text;
3065 if ( is_string( $text ) && !$this->incrementIncludeSize( 'post-expand', strlen( $text ) ) ) {
3066 # Error, oversize inclusion
3067 $text = "[[$originalTitle]]" .
3068 $this->insertStripItem( '<!-- WARNING: template omitted, post-expand include size too large -->' );
3069 $this->limitationWarn( 'post-expand-template-inclusion' );
3072 if ( $isLocalObj ) {
3073 $ret = array( 'object' => $text );
3075 $ret = array( 'text' => $text );
3078 wfProfileOut( __METHOD__
);
3083 * Get the semi-parsed DOM representation of a template with a given title,
3084 * and its redirect destination title. Cached.
3086 function getTemplateDom( $title ) {
3087 $cacheTitle = $title;
3088 $titleText = $title->getPrefixedDBkey();
3090 if ( isset( $this->mTplRedirCache
[$titleText] ) ) {
3091 list( $ns, $dbk ) = $this->mTplRedirCache
[$titleText];
3092 $title = Title
::makeTitle( $ns, $dbk );
3093 $titleText = $title->getPrefixedDBkey();
3095 if ( isset( $this->mTplDomCache
[$titleText] ) ) {
3096 return array( $this->mTplDomCache
[$titleText], $title );
3099 // Cache miss, go to the database
3100 list( $text, $title ) = $this->fetchTemplateAndTitle( $title );
3102 if ( $text === false ) {
3103 $this->mTplDomCache
[$titleText] = false;
3104 return array( false, $title );
3107 $dom = $this->preprocessToDom( $text, self
::PTD_FOR_INCLUSION
);
3108 $this->mTplDomCache
[ $titleText ] = $dom;
3110 if (! $title->equals($cacheTitle)) {
3111 $this->mTplRedirCache
[$cacheTitle->getPrefixedDBkey()] =
3112 array( $title->getNamespace(),$cdb = $title->getDBkey() );
3115 return array( $dom, $title );
3119 * Fetch the unparsed text of a template and register a reference to it.
3121 function fetchTemplateAndTitle( $title ) {
3122 $templateCb = $this->mOptions
->getTemplateCallback();
3123 $stuff = call_user_func( $templateCb, $title, $this );
3124 $text = $stuff['text'];
3125 $finalTitle = isset( $stuff['finalTitle'] ) ?
$stuff['finalTitle'] : $title;
3126 if ( isset( $stuff['deps'] ) ) {
3127 foreach ( $stuff['deps'] as $dep ) {
3128 $this->mOutput
->addTemplate( $dep['title'], $dep['page_id'], $dep['rev_id'] );
3131 return array($text,$finalTitle);
3134 function fetchTemplate( $title ) {
3135 $rv = $this->fetchTemplateAndTitle($title);
3140 * Static function to get a template
3141 * Can be overridden via ParserOptions::setTemplateCallback().
3143 static function statelessFetchTemplate( $title, $parser=false ) {
3144 $text = $skip = false;
3145 $finalTitle = $title;
3148 // Loop to fetch the article, with up to 1 redirect
3149 for ( $i = 0; $i < 2 && is_object( $title ); $i++
) {
3150 # Give extensions a chance to select the revision instead
3151 $id = false; // Assume current
3152 wfRunHooks( 'BeforeParserFetchTemplateAndtitle', array( $parser, &$title, &$skip, &$id ) );
3158 'page_id' => $title->getArticleID(),
3162 $rev = $id ? Revision
::newFromId( $id ) : Revision
::newFromTitle( $title );
3163 $rev_id = $rev ?
$rev->getId() : 0;
3164 // If there is no current revision, there is no page
3165 if( $id === false && !$rev ) {
3166 $linkCache = LinkCache
::singleton();
3167 $linkCache->addBadLinkObj( $title );
3172 'page_id' => $title->getArticleID(),
3173 'rev_id' => $rev_id );
3176 $text = $rev->getText();
3177 } elseif( $title->getNamespace() == NS_MEDIAWIKI
) {
3179 $message = $wgContLang->lcfirst( $title->getText() );
3180 $text = wfMsgForContentNoTrans( $message );
3181 if( wfEmptyMsg( $message, $text ) ) {
3188 if ( $text === false ) {
3192 $finalTitle = $title;
3193 $title = Title
::newFromRedirect( $text );
3197 'finalTitle' => $finalTitle,
3202 * Transclude an interwiki link.
3204 function interwikiTransclude( $title, $action ) {
3205 global $wgEnableScaryTranscluding;
3207 if (!$wgEnableScaryTranscluding)
3208 return wfMsg('scarytranscludedisabled');
3210 $url = $title->getFullUrl( "action=$action" );
3212 if (strlen($url) > 255)
3213 return wfMsg('scarytranscludetoolong');
3214 return $this->fetchScaryTemplateMaybeFromCache($url);
3217 function fetchScaryTemplateMaybeFromCache($url) {
3218 global $wgTranscludeCacheExpiry;
3219 $dbr = wfGetDB(DB_SLAVE
);
3220 $obj = $dbr->selectRow('transcache', array('tc_time', 'tc_contents'),
3221 array('tc_url' => $url));
3223 $time = $obj->tc_time
;
3224 $text = $obj->tc_contents
;
3225 if ($time && time() < $time +
$wgTranscludeCacheExpiry ) {
3230 $text = Http
::get($url);
3232 return wfMsg('scarytranscludefailed', $url);
3234 $dbw = wfGetDB(DB_MASTER
);
3235 $dbw->replace('transcache', array('tc_url'), array(
3237 'tc_time' => time(),
3238 'tc_contents' => $text));
3244 * Triple brace replacement -- used for template arguments
3247 function argSubstitution( $piece, $frame ) {
3248 wfProfileIn( __METHOD__
);
3251 $parts = $piece['parts'];
3252 $nameWithSpaces = $frame->expand( $piece['title'] );
3253 $argName = trim( $nameWithSpaces );
3255 $text = $frame->getArgument( $argName );
3256 if ( $text === false && $parts->getLength() > 0
3260 ||
( $this->ot
['wiki'] && $frame->isTemplate() )
3263 # No match in frame, use the supplied default
3264 $object = $parts->item( 0 )->getChildren();
3266 if ( !$this->incrementIncludeSize( 'arg', strlen( $text ) ) ) {
3267 $error = '<!-- WARNING: argument omitted, expansion size too large -->';
3268 $this->limitationWarn( 'post-expand-template-argument' );
3271 if ( $text === false && $object === false ) {
3273 $object = $frame->virtualBracketedImplode( '{{{', '|', '}}}', $nameWithSpaces, $parts );
3275 if ( $error !== false ) {
3278 if ( $object !== false ) {
3279 $ret = array( 'object' => $object );
3281 $ret = array( 'text' => $text );
3284 wfProfileOut( __METHOD__
);
3289 * Return the text to be used for a given extension tag.
3290 * This is the ghost of strip().
3292 * @param array $params Associative array of parameters:
3293 * name PPNode for the tag name
3294 * attr PPNode for unparsed text where tag attributes are thought to be
3295 * attributes Optional associative array of parsed attributes
3296 * inner Contents of extension element
3297 * noClose Original text did not have a close tag
3298 * @param PPFrame $frame
3300 function extensionSubstitution( $params, $frame ) {
3301 global $wgRawHtml, $wgContLang;
3303 $name = $frame->expand( $params['name'] );
3304 $attrText = !isset( $params['attr'] ) ?
null : $frame->expand( $params['attr'] );
3305 $content = !isset( $params['inner'] ) ?
null : $frame->expand( $params['inner'] );
3307 $marker = "{$this->mUniqPrefix}-$name-" . sprintf('%08X', $this->mMarkerIndex++
) . self
::MARKER_SUFFIX
;
3309 if ( $this->ot
['html'] ) {
3310 $name = strtolower( $name );
3312 $attributes = Sanitizer
::decodeTagAttributes( $attrText );
3313 if ( isset( $params['attributes'] ) ) {
3314 $attributes = $attributes +
$params['attributes'];
3322 throw new MWException( '<html> extension tag encountered unexpectedly' );
3325 $content = strtr($content, array('-{' => '-{', '}-' => '}-'));
3326 $output = Xml
::escapeTagsOnly( $content );
3329 $output = $wgContLang->armourMath(
3330 MathRenderer
::renderMath( $content, $attributes ) );
3333 $output = $this->renderImageGallery( $content, $attributes );
3336 if( isset( $this->mTagHooks
[$name] ) ) {
3337 # Workaround for PHP bug 35229 and similar
3338 if ( !is_callable( $this->mTagHooks
[$name] ) ) {
3339 throw new MWException( "Tag hook for $name is not callable\n" );
3341 $output = call_user_func_array( $this->mTagHooks
[$name],
3342 array( $content, $attributes, $this ) );
3344 $output = '<span class="error">Invalid tag extension name: ' .
3345 htmlspecialchars( $name ) . '</span>';
3349 if ( is_null( $attrText ) ) {
3352 if ( isset( $params['attributes'] ) ) {
3353 foreach ( $params['attributes'] as $attrName => $attrValue ) {
3354 $attrText .= ' ' . htmlspecialchars( $attrName ) . '="' .
3355 htmlspecialchars( $attrValue ) . '"';
3358 if ( $content === null ) {
3359 $output = "<$name$attrText/>";
3361 $close = is_null( $params['close'] ) ?
'' : $frame->expand( $params['close'] );
3362 $output = "<$name$attrText>$content$close";
3366 if ( $name === 'html' ||
$name === 'nowiki' ) {
3367 $this->mStripState
->nowiki
->setPair( $marker, $output );
3369 $this->mStripState
->general
->setPair( $marker, $output );
3375 * Increment an include size counter
3377 * @param string $type The type of expansion
3378 * @param integer $size The size of the text
3379 * @return boolean False if this inclusion would take it over the maximum, true otherwise
3381 function incrementIncludeSize( $type, $size ) {
3382 if ( $this->mIncludeSizes
[$type] +
$size > $this->mOptions
->getMaxIncludeSize( $type ) ) {
3385 $this->mIncludeSizes
[$type] +
= $size;
3391 * Increment the expensive function count
3393 * @return boolean False if the limit has been exceeded
3395 function incrementExpensiveFunctionCount() {
3396 global $wgExpensiveParserFunctionLimit;
3397 $this->mExpensiveFunctionCount++
;
3398 if($this->mExpensiveFunctionCount
<= $wgExpensiveParserFunctionLimit) {
3405 * Strip double-underscore items like __NOGALLERY__ and __NOTOC__
3406 * Fills $this->mDoubleUnderscores, returns the modified text
3408 function doDoubleUnderscore( $text ) {
3409 // The position of __TOC__ needs to be recorded
3410 $mw = MagicWord
::get( 'toc' );
3411 if( $mw->match( $text ) ) {
3412 $this->mShowToc
= true;
3413 $this->mForceTocPosition
= true;
3415 // Set a placeholder. At the end we'll fill it in with the TOC.
3416 $text = $mw->replace( '<!--MWTOC-->', $text, 1 );
3418 // Only keep the first one.
3419 $text = $mw->replace( '', $text );
3422 // Now match and remove the rest of them
3423 $mwa = MagicWord
::getDoubleUnderscoreArray();
3424 $this->mDoubleUnderscores
= $mwa->matchAndRemove( $text );
3426 if ( isset( $this->mDoubleUnderscores
['nogallery'] ) ) {
3427 $this->mOutput
->mNoGallery
= true;
3429 if ( isset( $this->mDoubleUnderscores
['notoc'] ) && !$this->mForceTocPosition
) {
3430 $this->mShowToc
= false;
3432 if ( isset( $this->mDoubleUnderscores
['hiddencat'] ) && $this->mTitle
->getNamespace() == NS_CATEGORY
) {
3433 $this->mOutput
->setProperty( 'hiddencat', 'y' );
3435 $containerCategory = Title
::makeTitleSafe( NS_CATEGORY
, wfMsgForContent( 'hidden-category-category' ) );
3436 if ( $containerCategory ) {
3437 $this->mOutput
->addCategory( $containerCategory->getDBkey(), $this->getDefaultSort() );
3439 wfDebug( __METHOD__
.": [[MediaWiki:hidden-category-category]] is not a valid title!\n" );
3442 if ( isset( $this->mDoubleUnderscores
['ignoreunused'] ) && $this->mTitle
->getNamespace() == NS_CATEGORY
) {
3443 $this->mOutput
->setProperty( 'ignoreunused', 'y' );
3445 # (bug 8068) Allow control over whether robots index a page.
3447 # FIXME (bug 14899): __INDEX__ always overrides __NOINDEX__ here! This
3448 # is not desirable, the last one on the page should win.
3449 if( isset( $this->mDoubleUnderscores
['noindex'] ) ) {
3450 $this->mOutput
->setIndexPolicy( 'noindex' );
3451 } elseif( isset( $this->mDoubleUnderscores
['index'] ) ) {
3452 $this->mOutput
->setIndexPolicy( 'index' );
3459 * This function accomplishes several tasks:
3460 * 1) Auto-number headings if that option is enabled
3461 * 2) Add an [edit] link to sections for users who have enabled the option and can edit the page
3462 * 3) Add a Table of contents on the top for users who have enabled the option
3463 * 4) Auto-anchor headings
3465 * It loops through all headlines, collects the necessary data, then splits up the
3466 * string and re-inserts the newly formatted headlines.
3468 * @param string $text
3469 * @param boolean $isMain
3472 function formatHeadings( $text, $isMain=true ) {
3473 global $wgMaxTocLevel, $wgContLang, $wgEnforceHtmlIds;
3475 $doNumberHeadings = $this->mOptions
->getNumberHeadings();
3476 $showEditLink = $this->mOptions
->getEditSection();
3478 // Do not call quickUserCan unless necessary
3479 if( $showEditLink && !$this->mTitle
->quickUserCan( 'edit' ) ) {
3483 # Inhibit editsection links if requested in the page
3484 if ( isset( $this->mDoubleUnderscores
['noeditsection'] ) ) {
3488 # Get all headlines for numbering them and adding funky stuff like [edit]
3489 # links - this is for later, but we need the number of headlines right now
3491 $numMatches = preg_match_all( '/<H(?P<level>[1-6])(?P<attrib>.*?'.'>)(?P<header>.*?)<\/H[1-6] *>/i', $text, $matches );
3493 # if there are fewer than 4 headlines in the article, do not show TOC
3494 # unless it's been explicitly enabled.
3495 $enoughToc = $this->mShowToc
&&
3496 (($numMatches >= 4) ||
$this->mForceTocPosition
);
3498 # Allow user to stipulate that a page should have a "new section"
3499 # link added via __NEWSECTIONLINK__
3500 if ( isset( $this->mDoubleUnderscores
['newsectionlink'] ) ) {
3501 $this->mOutput
->setNewSection( true );
3504 # if the string __FORCETOC__ (not case-sensitive) occurs in the HTML,
3505 # override above conditions and always show TOC above first header
3506 if ( isset( $this->mDoubleUnderscores
['forcetoc'] ) ) {
3507 $this->mShowToc
= true;
3511 # We need this to perform operations on the HTML
3512 $sk = $this->mOptions
->getSkin();
3518 # Ugh .. the TOC should have neat indentation levels which can be
3519 # passed to the skin functions. These are determined here
3523 $sublevelCount = array();
3524 $levelCount = array();
3530 $markerRegex = "{$this->mUniqPrefix}-h-(\d+)-" . self
::MARKER_SUFFIX
;
3531 $baseTitleText = $this->mTitle
->getPrefixedDBkey();
3534 foreach( $matches[3] as $headline ) {
3535 $isTemplate = false;
3537 $sectionIndex = false;
3539 $markerMatches = array();
3540 if (preg_match("/^$markerRegex/", $headline, $markerMatches)) {
3541 $serial = $markerMatches[1];
3542 list( $titleText, $sectionIndex ) = $this->mHeadings
[$serial];
3543 $isTemplate = ($titleText != $baseTitleText);
3544 $headline = preg_replace("/^$markerRegex/", "", $headline);
3548 $prevlevel = $level;
3549 $prevtoclevel = $toclevel;
3551 $level = $matches[1][$headlineCount];
3553 if( $doNumberHeadings ||
$enoughToc ) {
3555 if ( $level > $prevlevel ) {
3556 # Increase TOC level
3558 $sublevelCount[$toclevel] = 0;
3559 if( $toclevel<$wgMaxTocLevel ) {
3560 $prevtoclevel = $toclevel;
3561 $toc .= $sk->tocIndent();
3565 elseif ( $level < $prevlevel && $toclevel > 1 ) {
3566 # Decrease TOC level, find level to jump to
3568 if ( $toclevel == 2 && $level <= $levelCount[1] ) {
3569 # Can only go down to level 1
3572 for ($i = $toclevel; $i > 0; $i--) {
3573 if ( $levelCount[$i] == $level ) {
3574 # Found last matching level
3578 elseif ( $levelCount[$i] < $level ) {
3579 # Found first matching level below current level
3585 if( $toclevel<$wgMaxTocLevel ) {
3586 if($prevtoclevel < $wgMaxTocLevel) {
3587 # Unindent only if the previous toc level was shown :p
3588 $toc .= $sk->tocUnindent( $prevtoclevel - $toclevel );
3589 $prevtoclevel = $toclevel;
3591 $toc .= $sk->tocLineEnd();
3596 # No change in level, end TOC line
3597 if( $toclevel<$wgMaxTocLevel ) {
3598 $toc .= $sk->tocLineEnd();
3602 $levelCount[$toclevel] = $level;
3604 # count number of headlines for each level
3605 @$sublevelCount[$toclevel]++
;
3607 for( $i = 1; $i <= $toclevel; $i++
) {
3608 if( !empty( $sublevelCount[$i] ) ) {
3612 $numbering .= $wgContLang->formatNum( $sublevelCount[$i] );
3618 # The safe header is a version of the header text safe to use for links
3619 # Avoid insertion of weird stuff like <math> by expanding the relevant sections
3620 $safeHeadline = $this->mStripState
->unstripBoth( $headline );
3622 # Remove link placeholders by the link text.
3623 # <!--LINK number-->
3625 # link text with suffix
3626 $safeHeadline = $this->replaceLinkHoldersText( $safeHeadline );
3628 # Strip out HTML (other than plain <sup> and <sub>: bug 8393)
3629 $tocline = preg_replace(
3630 array( '#<(?!/?(sup|sub)).*?'.'>#', '#<(/?(sup|sub)).*?'.'>#' ),
3634 $tocline = trim( $tocline );
3636 # For the anchor, strip out HTML-y stuff period
3637 $safeHeadline = preg_replace( '/<.*?'.'>/', '', $safeHeadline );
3638 $safeHeadline = trim( $safeHeadline );
3640 # Save headline for section edit hint before it's escaped
3641 $headlineHint = $safeHeadline;
3643 if ( $wgEnforceHtmlIds ) {
3644 $legacyHeadline = false;
3645 $safeHeadline = Sanitizer
::escapeId( $safeHeadline,
3648 # For reverse compatibility, provide an id that's
3649 # HTML4-compatible, like we used to.
3651 # It may be worth noting, academically, that it's possible for
3652 # the legacy anchor to conflict with a non-legacy headline
3653 # anchor on the page. In this case likely the "correct" thing
3654 # would be to either drop the legacy anchors or make sure
3655 # they're numbered first. However, this would require people
3656 # to type in section names like "abc_.D7.93.D7.90.D7.A4"
3657 # manually, so let's not bother worrying about it.
3658 $legacyHeadline = Sanitizer
::escapeId( $safeHeadline,
3660 $safeHeadline = Sanitizer
::escapeId( $safeHeadline, 'xml' );
3662 if ( $legacyHeadline == $safeHeadline ) {
3663 # No reason to have both (in fact, we can't)
3664 $legacyHeadline = false;
3665 } elseif ( $legacyHeadline != Sanitizer
::escapeId(
3666 $legacyHeadline, 'xml' ) ) {
3667 # The legacy id is invalid XML. We used to allow this, but
3668 # there's no reason to do so anymore. Backward
3669 # compatibility will fail slightly in this case, but it's
3671 $legacyHeadline = false;
3675 # HTML names must be case-insensitively unique (bug 10721). FIXME:
3676 # Does this apply to Unicode characters? Because we aren't
3677 # handling those here.
3678 $arrayKey = strtolower( $safeHeadline );
3679 if ( $legacyHeadline === false ) {
3680 $legacyArrayKey = false;
3682 $legacyArrayKey = strtolower( $legacyHeadline );
3685 # count how many in assoc. array so we can track dupes in anchors
3686 if ( isset( $refers[$arrayKey] ) ) {
3687 $refers[$arrayKey]++
;
3689 $refers[$arrayKey] = 1;
3691 if ( isset( $refers[$legacyArrayKey] ) ) {
3692 $refers[$legacyArrayKey]++
;
3694 $refers[$legacyArrayKey] = 1;
3697 # Don't number the heading if it is the only one (looks silly)
3698 if( $doNumberHeadings && count( $matches[3] ) > 1) {
3699 # the two are different if the line contains a link
3700 $headline=$numbering . ' ' . $headline;
3703 # Create the anchor for linking from the TOC to the section
3704 $anchor = $safeHeadline;
3705 $legacyAnchor = $legacyHeadline;
3706 if ( $refers[$arrayKey] > 1 ) {
3707 $anchor .= '_' . $refers[$arrayKey];
3709 if ( $legacyHeadline !== false && $refers[$legacyArrayKey] > 1 ) {
3710 $legacyAnchor .= '_' . $refers[$legacyArrayKey];
3712 if( $enoughToc && ( !isset($wgMaxTocLevel) ||
$toclevel<$wgMaxTocLevel ) ) {
3713 $toc .= $sk->tocLine($anchor, $tocline, $numbering, $toclevel);
3714 $tocraw[] = array( 'toclevel' => $toclevel, 'level' => $level, 'line' => $tocline, 'number' => $numbering );
3716 # give headline the correct <h#> tag
3717 if( $showEditLink && $sectionIndex !== false ) {
3719 # Put a T flag in the section identifier, to indicate to extractSections()
3720 # that sections inside <includeonly> should be counted.
3721 $editlink = $sk->doEditSectionLink(Title
::newFromText( $titleText ), "T-$sectionIndex");
3723 $editlink = $sk->doEditSectionLink($this->mTitle
, $sectionIndex, $headlineHint);
3728 $head[$headlineCount] = $sk->makeHeadline( $level,
3729 $matches['attrib'][$headlineCount], $anchor, $headline,
3730 $editlink, $legacyAnchor );
3735 $this->mOutput
->setSections( $tocraw );
3737 # Never ever show TOC if no headers
3738 if( $numVisible < 1 ) {
3743 if( $prevtoclevel > 0 && $prevtoclevel < $wgMaxTocLevel ) {
3744 $toc .= $sk->tocUnindent( $prevtoclevel - 1 );
3746 $toc = $sk->tocList( $toc );
3749 # split up and insert constructed headlines
3751 $blocks = preg_split( '/<H[1-6].*?' . '>.*?<\/H[1-6]>/i', $text );
3754 foreach( $blocks as $block ) {
3755 if( $showEditLink && $headlineCount > 0 && $i == 0 && $block !== "\n" ) {
3756 # This is the [edit] link that appears for the top block of text when
3757 # section editing is enabled
3759 # Disabled because it broke block formatting
3760 # For example, a bullet point in the top line
3761 # $full .= $sk->editSectionLink(0);
3764 if( $enoughToc && !$i && $isMain && !$this->mForceTocPosition
) {
3765 # Top anchor now in skin
3769 if( !empty( $head[$i] ) ) {
3774 if( $this->mForceTocPosition
) {
3775 return str_replace( '<!--MWTOC-->', $toc, $full );
3782 * Transform wiki markup when saving a page by doing \r\n -> \n
3783 * conversion, substitting signatures, {{subst:}} templates, etc.
3785 * @param string $text the text to transform
3786 * @param Title &$title the Title object for the current article
3787 * @param User &$user the User object describing the current user
3788 * @param ParserOptions $options parsing options
3789 * @param bool $clearState whether to clear the parser state first
3790 * @return string the altered wiki markup
3793 function preSaveTransform( $text, &$title, $user, $options, $clearState = true ) {
3794 $this->mOptions
= $options;
3795 $this->setTitle( $title );
3796 $this->setOutputType( self
::OT_WIKI
);
3798 if ( $clearState ) {
3799 $this->clearState();
3805 $text = str_replace( array_keys( $pairs ), array_values( $pairs ), $text );
3806 $text = $this->pstPass2( $text, $user );
3807 $text = $this->mStripState
->unstripBoth( $text );
3812 * Pre-save transform helper function
3815 function pstPass2( $text, $user ) {
3816 global $wgContLang, $wgLocaltimezone;
3818 /* Note: This is the timestamp saved as hardcoded wikitext to
3819 * the database, we use $wgContLang here in order to give
3820 * everyone the same signature and use the default one rather
3821 * than the one selected in each user's preferences.
3823 * (see also bug 12815)
3825 $ts = $this->mOptions
->getTimestamp();
3826 $tz = wfMsgForContent( 'timezone-utc' );
3827 if ( isset( $wgLocaltimezone ) ) {
3828 $unixts = wfTimestamp( TS_UNIX
, $ts );
3829 $oldtz = getenv( 'TZ' );
3830 putenv( 'TZ='.$wgLocaltimezone );
3831 $ts = date( 'YmdHis', $unixts );
3832 $tz = date( 'T', $unixts ); # might vary on DST changeover!
3833 putenv( 'TZ='.$oldtz );
3836 $d = $wgContLang->timeanddate( $ts, false, false ) . " ($tz)";
3838 # Variable replacement
3839 # Because mOutputType is OT_WIKI, this will only process {{subst:xxx}} type tags
3840 $text = $this->replaceVariables( $text );
3843 $sigText = $this->getUserSig( $user );
3844 $text = strtr( $text, array(
3846 '~~~~' => "$sigText $d",
3850 # Context links: [[|name]] and [[name (context)|]]
3852 global $wgLegalTitleChars;
3853 $tc = "[$wgLegalTitleChars]";
3854 $nc = '[ _0-9A-Za-z\x80-\xff-]'; # Namespaces can use non-ascii!
3856 $p1 = "/\[\[(:?$nc+:|:|)($tc+?)( \\($tc+\\))\\|]]/"; # [[ns:page (context)|]]
3857 $p4 = "/\[\[(:?$nc+:|:|)($tc+?)(($tc+))\\|]]/"; # [[ns:page(context)|]]
3858 $p3 = "/\[\[(:?$nc+:|:|)($tc+?)( \\($tc+\\)|)(, $tc+|)\\|]]/"; # [[ns:page (context), context|]]
3859 $p2 = "/\[\[\\|($tc+)]]/"; # [[|page]]
3861 # try $p1 first, to turn "[[A, B (C)|]]" into "[[A, B (C)|A, B]]"
3862 $text = preg_replace( $p1, '[[\\1\\2\\3|\\2]]', $text );
3863 $text = preg_replace( $p4, '[[\\1\\2\\3|\\2]]', $text );
3864 $text = preg_replace( $p3, '[[\\1\\2\\3\\4|\\2]]', $text );
3866 $t = $this->mTitle
->getText();
3868 if ( preg_match( "/^($nc+:|)$tc+?( \\($tc+\\))$/", $t, $m ) ) {
3869 $text = preg_replace( $p2, "[[$m[1]\\1$m[2]|\\1]]", $text );
3870 } elseif ( preg_match( "/^($nc+:|)$tc+?(, $tc+|)$/", $t, $m ) && '' != "$m[1]$m[2]" ) {
3871 $text = preg_replace( $p2, "[[$m[1]\\1$m[2]|\\1]]", $text );
3873 # if there's no context, don't bother duplicating the title
3874 $text = preg_replace( $p2, '[[\\1]]', $text );
3877 # Trim trailing whitespace
3878 $text = rtrim( $text );
3884 * Fetch the user's signature text, if any, and normalize to
3885 * validated, ready-to-insert wikitext.
3891 function getUserSig( &$user ) {
3892 global $wgMaxSigChars;
3894 $username = $user->getName();
3895 $nickname = $user->getOption( 'nickname' );
3896 $nickname = $nickname === '' ?
$username : $nickname;
3898 if( mb_strlen( $nickname ) > $wgMaxSigChars ) {
3899 $nickname = $username;
3900 wfDebug( __METHOD__
. ": $username has overlong signature.\n" );
3901 } elseif( $user->getBoolOption( 'fancysig' ) !== false ) {
3902 # Sig. might contain markup; validate this
3903 if( $this->validateSig( $nickname ) !== false ) {
3904 # Validated; clean up (if needed) and return it
3905 return $this->cleanSig( $nickname, true );
3907 # Failed to validate; fall back to the default
3908 $nickname = $username;
3909 wfDebug( __METHOD__
.": $username has bad XML tags in signature.\n" );
3913 // Make sure nickname doesnt get a sig in a sig
3914 $nickname = $this->cleanSigInSig( $nickname );
3916 # If we're still here, make it a link to the user page
3917 $userText = wfEscapeWikiText( $username );
3918 $nickText = wfEscapeWikiText( $nickname );
3919 if ( $user->isAnon() ) {
3920 return wfMsgExt( 'signature-anon', array( 'content', 'parsemag' ), $userText, $nickText );
3922 return wfMsgExt( 'signature', array( 'content', 'parsemag' ), $userText, $nickText );
3927 * Check that the user's signature contains no bad XML
3929 * @param string $text
3930 * @return mixed An expanded string, or false if invalid.
3932 function validateSig( $text ) {
3933 return( Xml
::isWellFormedXmlFragment( $text ) ?
$text : false );
3937 * Clean up signature text
3939 * 1) Strip ~~~, ~~~~ and ~~~~~ out of signatures @see cleanSigInSig
3940 * 2) Substitute all transclusions
3942 * @param string $text
3943 * @param $parsing Whether we're cleaning (preferences save) or parsing
3944 * @return string Signature text
3946 function cleanSig( $text, $parsing = false ) {
3949 $this->clearState();
3950 $this->setTitle( $wgTitle );
3951 $this->mOptions
= new ParserOptions
;
3952 $this->setOutputType
= self
::OT_PREPROCESS
;
3955 # Option to disable this feature
3956 if ( !$this->mOptions
->getCleanSignatures() ) {
3960 # FIXME: regex doesn't respect extension tags or nowiki
3961 # => Move this logic to braceSubstitution()
3962 $substWord = MagicWord
::get( 'subst' );
3963 $substRegex = '/\{\{(?!(?:' . $substWord->getBaseRegex() . '))/x' . $substWord->getRegexCase();
3964 $substText = '{{' . $substWord->getSynonym( 0 );
3966 $text = preg_replace( $substRegex, $substText, $text );
3967 $text = $this->cleanSigInSig( $text );
3968 $dom = $this->preprocessToDom( $text );
3969 $frame = $this->getPreprocessor()->newFrame();
3970 $text = $frame->expand( $dom );
3973 $text = $this->mStripState
->unstripBoth( $text );
3980 * Strip ~~~, ~~~~ and ~~~~~ out of signatures
3981 * @param string $text
3982 * @return string Signature text with /~{3,5}/ removed
3984 function cleanSigInSig( $text ) {
3985 $text = preg_replace( '/~{3,5}/', '', $text );
3990 * Set up some variables which are usually set up in parse()
3991 * so that an external function can call some class members with confidence
3994 function startExternalParse( &$title, $options, $outputType, $clearState = true ) {
3995 $this->setTitle( $title );
3996 $this->mOptions
= $options;
3997 $this->setOutputType( $outputType );
3998 if ( $clearState ) {
3999 $this->clearState();
4004 * Wrapper for preprocess()
4006 * @param string $text the text to preprocess
4007 * @param ParserOptions $options options
4011 function transformMsg( $text, $options ) {
4013 static $executing = false;
4015 # Guard against infinite recursion
4021 wfProfileIn(__METHOD__
);
4022 $text = $this->preprocess( $text, $wgTitle, $options );
4025 wfProfileOut(__METHOD__
);
4030 * Create an HTML-style tag, e.g. <yourtag>special text</yourtag>
4031 * The callback should have the following form:
4032 * function myParserHook( $text, $params, &$parser ) { ... }
4034 * Transform and return $text. Use $parser for any required context, e.g. use
4035 * $parser->getTitle() and $parser->getOptions() not $wgTitle or $wgOut->mParserOptions
4039 * @param mixed $tag The tag to use, e.g. 'hook' for <hook>
4040 * @param mixed $callback The callback function (and object) to use for the tag
4042 * @return The old value of the mTagHooks array associated with the hook
4044 function setHook( $tag, $callback ) {
4045 $tag = strtolower( $tag );
4046 $oldVal = isset( $this->mTagHooks
[$tag] ) ?
$this->mTagHooks
[$tag] : null;
4047 $this->mTagHooks
[$tag] = $callback;
4048 if( !in_array( $tag, $this->mStripList
) ) {
4049 $this->mStripList
[] = $tag;
4055 function setTransparentTagHook( $tag, $callback ) {
4056 $tag = strtolower( $tag );
4057 $oldVal = isset( $this->mTransparentTagHooks
[$tag] ) ?
$this->mTransparentTagHooks
[$tag] : null;
4058 $this->mTransparentTagHooks
[$tag] = $callback;
4064 * Remove all tag hooks
4066 function clearTagHooks() {
4067 $this->mTagHooks
= array();
4068 $this->mStripList
= $this->mDefaultStripList
;
4072 * Create a function, e.g. {{sum:1|2|3}}
4073 * The callback function should have the form:
4074 * function myParserFunction( &$parser, $arg1, $arg2, $arg3 ) { ... }
4076 * Or with SFH_OBJECT_ARGS:
4077 * function myParserFunction( $parser, $frame, $args ) { ... }
4079 * The callback may either return the text result of the function, or an array with the text
4080 * in element 0, and a number of flags in the other elements. The names of the flags are
4081 * specified in the keys. Valid flags are:
4082 * found The text returned is valid, stop processing the template. This
4084 * nowiki Wiki markup in the return value should be escaped
4085 * isHTML The returned text is HTML, armour it against wikitext transformation
4089 * @param string $id The magic word ID
4090 * @param mixed $callback The callback function (and object) to use
4091 * @param integer $flags a combination of the following flags:
4092 * SFH_NO_HASH No leading hash, i.e. {{plural:...}} instead of {{#if:...}}
4094 * SFH_OBJECT_ARGS Pass the template arguments as PPNode objects instead of text. This
4095 * allows for conditional expansion of the parse tree, allowing you to eliminate dead
4096 * branches and thus speed up parsing. It is also possible to analyse the parse tree of
4097 * the arguments, and to control the way they are expanded.
4099 * The $frame parameter is a PPFrame. This can be used to produce expanded text from the
4100 * arguments, for instance:
4101 * $text = isset( $args[0] ) ? $frame->expand( $args[0] ) : '';
4103 * For technical reasons, $args[0] is pre-expanded and will be a string. This may change in
4104 * future versions. Please call $frame->expand() on it anyway so that your code keeps
4105 * working if/when this is changed.
4107 * If you want whitespace to be trimmed from $args, you need to do it yourself, post-
4110 * Please read the documentation in includes/parser/Preprocessor.php for more information
4111 * about the methods available in PPFrame and PPNode.
4113 * @return The old callback function for this name, if any
4115 function setFunctionHook( $id, $callback, $flags = 0 ) {
4116 $oldVal = isset( $this->mFunctionHooks
[$id] ) ?
$this->mFunctionHooks
[$id][0] : null;
4117 $this->mFunctionHooks
[$id] = array( $callback, $flags );
4119 # Add to function cache
4120 $mw = MagicWord
::get( $id );
4122 throw new MWException( __METHOD__
.'() expecting a magic word identifier.' );
4124 $synonyms = $mw->getSynonyms();
4125 $sensitive = intval( $mw->isCaseSensitive() );
4127 foreach ( $synonyms as $syn ) {
4129 if ( !$sensitive ) {
4130 $syn = strtolower( $syn );
4133 if ( !( $flags & SFH_NO_HASH
) ) {
4136 # Remove trailing colon
4137 if ( substr( $syn, -1, 1 ) === ':' ) {
4138 $syn = substr( $syn, 0, -1 );
4140 $this->mFunctionSynonyms
[$sensitive][$syn] = $id;
4146 * Get all registered function hook identifiers
4150 function getFunctionHooks() {
4151 return array_keys( $this->mFunctionHooks
);
4155 * Replace <!--LINK--> link placeholders with actual links, in the buffer
4156 * Placeholders created in Skin::makeLinkObj()
4157 * Returns an array of link CSS classes, indexed by PDBK.
4159 function replaceLinkHolders( &$text, $options = 0 ) {
4160 return $this->mLinkHolders
->replace( $text );
4164 * Replace <!--LINK--> link placeholders with plain text of links
4165 * (not HTML-formatted).
4166 * @param string $text
4169 function replaceLinkHoldersText( $text ) {
4170 return $this->mLinkHolders
->replaceText( $text );
4174 * Tag hook handler for 'pre'.
4176 function renderPreTag( $text, $attribs ) {
4177 // Backwards-compatibility hack
4178 $content = StringUtils
::delimiterReplace( '<nowiki>', '</nowiki>', '$1', $text, 'i' );
4180 $attribs = Sanitizer
::validateTagAttributes( $attribs, 'pre' );
4181 return Xml
::openElement( 'pre', $attribs ) .
4182 Xml
::escapeTagsOnly( $content ) .
4187 * Renders an image gallery from a text with one line per image.
4188 * text labels may be given by using |-style alternative text. E.g.
4189 * Image:one.jpg|The number "1"
4190 * Image:tree.jpg|A tree
4191 * given as text will return the HTML of a gallery with two images,
4192 * labeled 'The number "1"' and
4195 function renderImageGallery( $text, $params ) {
4196 $ig = new ImageGallery();
4197 $ig->setContextTitle( $this->mTitle
);
4198 $ig->setShowBytes( false );
4199 $ig->setShowFilename( false );
4200 $ig->setParser( $this );
4201 $ig->setHideBadImages();
4202 $ig->setAttributes( Sanitizer
::validateTagAttributes( $params, 'table' ) );
4203 $ig->useSkin( $this->mOptions
->getSkin() );
4204 $ig->mRevisionId
= $this->mRevisionId
;
4206 if( isset( $params['caption'] ) ) {
4207 $caption = $params['caption'];
4208 $caption = htmlspecialchars( $caption );
4209 $caption = $this->replaceInternalLinks( $caption );
4210 $ig->setCaptionHtml( $caption );
4212 if( isset( $params['perrow'] ) ) {
4213 $ig->setPerRow( $params['perrow'] );
4215 if( isset( $params['widths'] ) ) {
4216 $ig->setWidths( $params['widths'] );
4218 if( isset( $params['heights'] ) ) {
4219 $ig->setHeights( $params['heights'] );
4222 wfRunHooks( 'BeforeParserrenderImageGallery', array( &$this, &$ig ) );
4224 $lines = StringUtils
::explode( "\n", $text );
4225 foreach ( $lines as $line ) {
4226 # match lines like these:
4227 # Image:someimage.jpg|This is some image
4229 preg_match( "/^([^|]+)(\\|(.*))?$/", $line, $matches );
4231 if ( count( $matches ) == 0 ) {
4235 if ( strpos( $matches[0], '%' ) !== false )
4236 $matches[1] = urldecode( $matches[1] );
4237 $tp = Title
::newFromText( $matches[1]/*, NS_FILE*/ );
4239 if( is_null( $nt ) ) {
4240 # Bogus title. Ignore these so we don't bomb out later.
4243 if ( isset( $matches[3] ) ) {
4244 $label = $matches[3];
4249 $html = $this->recursiveTagParse( trim( $label ) );
4251 $ig->add( $nt, $html );
4253 # Only add real images (bug #5586)
4254 if ( $nt->getNamespace() == NS_FILE
) {
4255 $this->mOutput
->addImage( $nt->getDBkey() );
4258 return $ig->toHTML();
4261 function getImageParams( $handler ) {
4263 $handlerClass = get_class( $handler );
4267 if ( !isset( $this->mImageParams
[$handlerClass] ) ) {
4268 // Initialise static lists
4269 static $internalParamNames = array(
4270 'horizAlign' => array( 'left', 'right', 'center', 'none' ),
4271 'vertAlign' => array( 'baseline', 'sub', 'super', 'top', 'text-top', 'middle',
4272 'bottom', 'text-bottom' ),
4273 'frame' => array( 'thumbnail', 'manualthumb', 'framed', 'frameless',
4274 'upright', 'border', 'link', 'alt' ),
4276 static $internalParamMap;
4277 if ( !$internalParamMap ) {
4278 $internalParamMap = array();
4279 foreach ( $internalParamNames as $type => $names ) {
4280 foreach ( $names as $name ) {
4281 $magicName = str_replace( '-', '_', "img_$name" );
4282 $internalParamMap[$magicName] = array( $type, $name );
4287 // Add handler params
4288 $paramMap = $internalParamMap;
4290 $handlerParamMap = $handler->getParamMap();
4291 foreach ( $handlerParamMap as $magic => $paramName ) {
4292 $paramMap[$magic] = array( 'handler', $paramName );
4295 $this->mImageParams
[$handlerClass] = $paramMap;
4296 $this->mImageParamsMagicArray
[$handlerClass] = new MagicWordArray( array_keys( $paramMap ) );
4298 return array( $this->mImageParams
[$handlerClass], $this->mImageParamsMagicArray
[$handlerClass] );
4302 * Parse image options text and use it to make an image
4303 * @param Title $title
4304 * @param string $options
4305 * @param LinkHolderArray $holders
4307 function makeImage( $title, $options, $holders = false ) {
4308 # Check if the options text is of the form "options|alt text"
4310 # * thumbnail make a thumbnail with enlarge-icon and caption, alignment depends on lang
4311 # * left no resizing, just left align. label is used for alt= only
4312 # * right same, but right aligned
4313 # * none same, but not aligned
4314 # * ___px scale to ___ pixels width, no aligning. e.g. use in taxobox
4315 # * center center the image
4316 # * framed Keep original image size, no magnify-button.
4317 # * frameless like 'thumb' but without a frame. Keeps user preferences for width
4318 # * upright reduce width for upright images, rounded to full __0 px
4319 # * border draw a 1px border around the image
4320 # * alt Text for HTML alt attribute (defaults to empty)
4321 # vertical-align values (no % or length right now):
4331 $parts = StringUtils
::explode( "|", $options );
4332 $sk = $this->mOptions
->getSkin();
4334 # Give extensions a chance to select the file revision for us
4335 $skip = $time = $descQuery = false;
4336 wfRunHooks( 'BeforeParserMakeImageLinkObj', array( &$this, &$title, &$skip, &$time, &$descQuery ) );
4339 return $sk->link( $title );
4343 $imagename = $title->getDBkey();
4344 if ( isset( $this->mFileCache
[$imagename][$time] ) ) {
4345 $file = $this->mFileCache
[$imagename][$time];
4347 $file = wfFindFile( $title, $time );
4348 if ( count( $this->mFileCache
) > 1000 ) {
4349 $this->mFileCache
= array();
4351 $this->mFileCache
[$imagename][$time] = $file;
4354 $handler = $file ?
$file->getHandler() : false;
4356 list( $paramMap, $mwArray ) = $this->getImageParams( $handler );
4358 # Process the input parameters
4360 $params = array( 'frame' => array(), 'handler' => array(),
4361 'horizAlign' => array(), 'vertAlign' => array() );
4362 foreach( $parts as $part ) {
4363 $part = trim( $part );
4364 list( $magicName, $value ) = $mwArray->matchVariableStartToEnd( $part );
4366 if( isset( $paramMap[$magicName] ) ) {
4367 list( $type, $paramName ) = $paramMap[$magicName];
4369 // Special case; width and height come in one variable together
4370 if( $type === 'handler' && $paramName === 'width' ) {
4372 # (bug 13500) In both cases (width/height and width only),
4373 # permit trailing "px" for backward compatibility.
4374 if ( preg_match( '/^([0-9]*)x([0-9]*)\s*(?:px)?\s*$/', $value, $m ) ) {
4375 $width = intval( $m[1] );
4376 $height = intval( $m[2] );
4377 if ( $handler->validateParam( 'width', $width ) ) {
4378 $params[$type]['width'] = $width;
4381 if ( $handler->validateParam( 'height', $height ) ) {
4382 $params[$type]['height'] = $height;
4385 } elseif ( preg_match( '/^[0-9]*\s*(?:px)?\s*$/', $value ) ) {
4386 $width = intval( $value );
4387 if ( $handler->validateParam( 'width', $width ) ) {
4388 $params[$type]['width'] = $width;
4391 } // else no validation -- bug 13436
4393 if ( $type === 'handler' ) {
4394 # Validate handler parameter
4395 $validated = $handler->validateParam( $paramName, $value );
4397 # Validate internal parameters
4398 switch( $paramName ) {
4401 // @fixme - possibly check validity here for
4402 // manualthumb? downstream behavior seems odd with
4403 // missing manual thumbs.
4405 $value = $this->stripAltText( $value, $holders );
4408 $chars = self
::EXT_LINK_URL_CLASS
;
4409 $prots = $this->mUrlProtocols
;
4410 if ( $value === '' ) {
4411 $paramName = 'no-link';
4414 } elseif ( preg_match( "/^$prots/", $value ) ) {
4415 if ( preg_match( "/^($prots)$chars+$/", $value, $m ) ) {
4416 $paramName = 'link-url';
4417 $this->mOutput
->addExternalLink( $value );
4421 $linkTitle = Title
::newFromText( $value );
4423 $paramName = 'link-title';
4424 $value = $linkTitle;
4425 $this->mOutput
->addLink( $linkTitle );
4431 // Most other things appear to be empty or numeric...
4432 $validated = ( $value === false ||
is_numeric( trim( $value ) ) );
4437 $params[$type][$paramName] = $value;
4441 if ( !$validated ) {
4446 # Process alignment parameters
4447 if ( $params['horizAlign'] ) {
4448 $params['frame']['align'] = key( $params['horizAlign'] );
4450 if ( $params['vertAlign'] ) {
4451 $params['frame']['valign'] = key( $params['vertAlign'] );
4454 $params['frame']['caption'] = $caption;
4456 $params['frame']['title'] = $this->stripAltText( $caption, $holders );
4458 # In the old days, [[Image:Foo|text...]] would set alt text. Later it
4459 # came to also set the caption, ordinary text after the image -- which
4460 # makes no sense, because that just repeats the text multiple times in
4461 # screen readers. It *also* came to set the title attribute.
4463 # Now that we have an alt attribute, we should not set the alt text to
4464 # equal the caption: that's worse than useless, it just repeats the
4465 # text. This is the framed/thumbnail case. If there's no caption, we
4466 # use the unnamed parameter for alt text as well, just for the time be-
4467 # ing, if the unnamed param is set and the alt param is not.
4469 # For the future, we need to figure out if we want to tweak this more,
4470 # e.g., introducing a title= parameter for the title; ignoring the un-
4471 # named parameter entirely for images without a caption; adding an ex-
4472 # plicit caption= parameter and preserving the old magic unnamed para-
4474 if( $caption !== '' && !isset( $params['frame']['alt'] )
4475 && !isset( $params['frame']['framed'] )
4476 && !isset( $params['frame']['thumbnail'] )
4477 && !isset( $params['frame']['manualthumb'] ) ) {
4478 $params['frame']['alt'] = $params['frame']['title'];
4481 wfRunHooks( 'ParserMakeImageParams', array( $title, $file, &$params ) );
4483 # Linker does the rest
4484 $ret = $sk->makeImageLink2( $title, $file, $params['frame'], $params['handler'], $time, $descQuery );
4486 # Give the handler a chance to modify the parser object
4488 $handler->parserTransformHook( $this, $file );
4494 protected function stripAltText( $caption, $holders ) {
4495 # Strip bad stuff out of the title (tooltip). We can't just use
4496 # replaceLinkHoldersText() here, because if this function is called
4497 # from replaceInternalLinks2(), mLinkHolders won't be up-to-date.
4499 $tooltip = $holders->replaceText( $caption );
4501 $tooltip = $this->replaceLinkHoldersText( $caption );
4504 # make sure there are no placeholders in thumbnail attributes
4505 # that are later expanded to html- so expand them now and
4507 $tooltip = $this->mStripState
->unstripBoth( $tooltip );
4508 $tooltip = Sanitizer
::stripAllTags( $tooltip );
4514 * Set a flag in the output object indicating that the content is dynamic and
4515 * shouldn't be cached.
4517 function disableCache() {
4518 wfDebug( "Parser output marked as uncacheable.\n" );
4519 $this->mOutput
->mCacheTime
= -1;
4523 * Callback from the Sanitizer for expanding items found in HTML attribute
4524 * values, so they can be safely tested and escaped.
4525 * @param string $text
4526 * @param PPFrame $frame
4530 function attributeStripCallback( &$text, $frame = false ) {
4531 $text = $this->replaceVariables( $text, $frame );
4532 $text = $this->mStripState
->unstripBoth( $text );
4541 function Title( $x = NULL ) { return wfSetVar( $this->mTitle
, $x ); }
4542 function Options( $x = NULL ) { return wfSetVar( $this->mOptions
, $x ); }
4543 function OutputType( $x = NULL ) { return wfSetVar( $this->mOutputType
, $x ); }
4549 function getTags() { return array_merge( array_keys($this->mTransparentTagHooks
), array_keys( $this->mTagHooks
) ); }
4554 * Break wikitext input into sections, and either pull or replace
4555 * some particular section's text.
4557 * External callers should use the getSection and replaceSection methods.
4559 * @param string $text Page wikitext
4560 * @param string $section A section identifier string of the form:
4561 * <flag1> - <flag2> - ... - <section number>
4563 * Currently the only recognised flag is "T", which means the target section number
4564 * was derived during a template inclusion parse, in other words this is a template
4565 * section edit link. If no flags are given, it was an ordinary section edit link.
4566 * This flag is required to avoid a section numbering mismatch when a section is
4567 * enclosed by <includeonly> (bug 6563).
4569 * The section number 0 pulls the text before the first heading; other numbers will
4570 * pull the given section along with its lower-level subsections. If the section is
4571 * not found, $mode=get will return $newtext, and $mode=replace will return $text.
4573 * @param string $mode One of "get" or "replace"
4574 * @param string $newText Replacement text for section data.
4575 * @return string for "get", the extracted section text.
4576 * for "replace", the whole page with the section replaced.
4578 private function extractSections( $text, $section, $mode, $newText='' ) {
4580 $this->clearState();
4581 $this->setTitle( $wgTitle ); // not generally used but removes an ugly failure mode
4582 $this->mOptions
= new ParserOptions
;
4583 $this->setOutputType( self
::OT_WIKI
);
4585 $frame = $this->getPreprocessor()->newFrame();
4587 // Process section extraction flags
4589 $sectionParts = explode( '-', $section );
4590 $sectionIndex = array_pop( $sectionParts );
4591 foreach ( $sectionParts as $part ) {
4592 if ( $part === 'T' ) {
4593 $flags |
= self
::PTD_FOR_INCLUSION
;
4596 // Preprocess the text
4597 $root = $this->preprocessToDom( $text, $flags );
4599 // <h> nodes indicate section breaks
4600 // They can only occur at the top level, so we can find them by iterating the root's children
4601 $node = $root->getFirstChild();
4603 // Find the target section
4604 if ( $sectionIndex == 0 ) {
4605 // Section zero doesn't nest, level=big
4606 $targetLevel = 1000;
4609 if ( $node->getName() === 'h' ) {
4610 $bits = $node->splitHeading();
4611 if ( $bits['i'] == $sectionIndex ) {
4612 $targetLevel = $bits['level'];
4616 if ( $mode === 'replace' ) {
4617 $outText .= $frame->expand( $node, PPFrame
::RECOVER_ORIG
);
4619 $node = $node->getNextSibling();
4625 if ( $mode === 'get' ) {
4632 // Find the end of the section, including nested sections
4634 if ( $node->getName() === 'h' ) {
4635 $bits = $node->splitHeading();
4636 $curLevel = $bits['level'];
4637 if ( $bits['i'] != $sectionIndex && $curLevel <= $targetLevel ) {
4641 if ( $mode === 'get' ) {
4642 $outText .= $frame->expand( $node, PPFrame
::RECOVER_ORIG
);
4644 $node = $node->getNextSibling();
4647 // Write out the remainder (in replace mode only)
4648 if ( $mode === 'replace' ) {
4649 // Output the replacement text
4650 // Add two newlines on -- trailing whitespace in $newText is conventionally
4651 // stripped by the editor, so we need both newlines to restore the paragraph gap
4652 $outText .= $newText . "\n\n";
4654 $outText .= $frame->expand( $node, PPFrame
::RECOVER_ORIG
);
4655 $node = $node->getNextSibling();
4659 if ( is_string( $outText ) ) {
4660 // Re-insert stripped tags
4661 $outText = rtrim( $this->mStripState
->unstripBoth( $outText ) );
4668 * This function returns the text of a section, specified by a number ($section).
4669 * A section is text under a heading like == Heading == or \<h1\>Heading\</h1\>, or
4670 * the first section before any such heading (section 0).
4672 * If a section contains subsections, these are also returned.
4674 * @param string $text text to look in
4675 * @param string $section section identifier
4676 * @param string $deftext default to return if section is not found
4677 * @return string text of the requested section
4679 public function getSection( $text, $section, $deftext='' ) {
4680 return $this->extractSections( $text, $section, "get", $deftext );
4683 public function replaceSection( $oldtext, $section, $text ) {
4684 return $this->extractSections( $oldtext, $section, "replace", $text );
4688 * Get the timestamp associated with the current revision, adjusted for
4689 * the default server-local timestamp
4691 function getRevisionTimestamp() {
4692 if ( is_null( $this->mRevisionTimestamp
) ) {
4693 wfProfileIn( __METHOD__
);
4695 $dbr = wfGetDB( DB_SLAVE
);
4696 $timestamp = $dbr->selectField( 'revision', 'rev_timestamp',
4697 array( 'rev_id' => $this->mRevisionId
), __METHOD__
);
4699 // Normalize timestamp to internal MW format for timezone processing.
4700 // This has the added side-effect of replacing a null value with
4701 // the current time, which gives us more sensible behavior for
4703 $timestamp = wfTimestamp( TS_MW
, $timestamp );
4705 // The cryptic '' timezone parameter tells to use the site-default
4706 // timezone offset instead of the user settings.
4708 // Since this value will be saved into the parser cache, served
4709 // to other users, and potentially even used inside links and such,
4710 // it needs to be consistent for all visitors.
4711 $this->mRevisionTimestamp
= $wgContLang->userAdjust( $timestamp, '' );
4713 wfProfileOut( __METHOD__
);
4715 return $this->mRevisionTimestamp
;
4719 * Mutator for $mDefaultSort
4721 * @param $sort New value
4723 public function setDefaultSort( $sort ) {
4724 $this->mDefaultSort
= $sort;
4728 * Accessor for $mDefaultSort
4729 * Will use the title/prefixed title if none is set
4733 public function getDefaultSort() {
4734 global $wgCategoryPrefixedDefaultSortkey;
4735 if( $this->mDefaultSort
!== false ) {
4736 return $this->mDefaultSort
;
4737 } elseif ($this->mTitle
->getNamespace() == NS_CATEGORY ||
4738 !$wgCategoryPrefixedDefaultSortkey) {
4739 return $this->mTitle
->getText();
4741 return $this->mTitle
->getPrefixedText();
4746 * Accessor for $mDefaultSort
4747 * Unlike getDefaultSort(), will return false if none is set
4749 * @return string or false
4751 public function getCustomDefaultSort() {
4752 return $this->mDefaultSort
;
4756 * Try to guess the section anchor name based on a wikitext fragment
4757 * presumably extracted from a heading, for example "Header" from
4760 public function guessSectionNameFromWikiText( $text ) {
4761 # Strip out wikitext links(they break the anchor)
4762 $text = $this->stripSectionName( $text );
4763 $headline = Sanitizer
::decodeCharReferences( $text );
4765 $headline = StringUtils
::delimiterReplace( '<', '>', '', $headline );
4766 $headline = trim( $headline );
4767 $sectionanchor = '#' . urlencode( str_replace( ' ', '_', $headline ) );
4768 $replacearray = array(
4773 array_keys( $replacearray ),
4774 array_values( $replacearray ),
4779 * Strips a text string of wikitext for use in a section anchor
4781 * Accepts a text string and then removes all wikitext from the
4782 * string and leaves only the resultant text (i.e. the result of
4783 * [[User:WikiSysop|Sysop]] would be "Sysop" and the result of
4784 * [[User:WikiSysop]] would be "User:WikiSysop") - this is intended
4785 * to create valid section anchors by mimicing the output of the
4786 * parser when headings are parsed.
4788 * @param $text string Text string to be stripped of wikitext
4789 * for use in a Section anchor
4790 * @return Filtered text string
4792 public function stripSectionName( $text ) {
4793 # Strip internal link markup
4794 $text = preg_replace('/\[\[:?([^[|]+)\|([^[]+)\]\]/','$2',$text);
4795 $text = preg_replace('/\[\[:?([^[]+)\|?\]\]/','$1',$text);
4797 # Strip external link markup (FIXME: Not Tolerant to blank link text
4798 # I.E. [http://www.mediawiki.org] will render as [1] or something depending
4799 # on how many empty links there are on the page - need to figure that out.
4800 $text = preg_replace('/\[(?:' . wfUrlProtocols() . ')([^ ]+?) ([^[]+)\]/','$2',$text);
4802 # Parse wikitext quotes (italics & bold)
4803 $text = $this->doQuotes($text);
4806 $text = StringUtils
::delimiterReplace( '<', '>', '', $text );
4810 function srvus( $text ) {
4811 return $this->testSrvus( $text, $this->mOutputType
);
4815 * strip/replaceVariables/unstrip for preprocessor regression testing
4817 function testSrvus( $text, $title, $options, $outputType = self
::OT_HTML
) {
4818 $this->clearState();
4819 if ( ! ( $title instanceof Title
) ) {
4820 $title = Title
::newFromText( $title );
4822 $this->mTitle
= $title;
4823 $this->mOptions
= $options;
4824 $this->setOutputType( $outputType );
4825 $text = $this->replaceVariables( $text );
4826 $text = $this->mStripState
->unstripBoth( $text );
4827 $text = Sanitizer
::removeHTMLtags( $text );
4831 function testPst( $text, $title, $options ) {
4833 if ( ! ( $title instanceof Title
) ) {
4834 $title = Title
::newFromText( $title );
4836 return $this->preSaveTransform( $text, $title, $wgUser, $options );
4839 function testPreprocess( $text, $title, $options ) {
4840 if ( ! ( $title instanceof Title
) ) {
4841 $title = Title
::newFromText( $title );
4843 return $this->testSrvus( $text, $title, $options, self
::OT_PREPROCESS
);
4846 function markerSkipCallback( $s, $callback ) {
4849 while ( $i < strlen( $s ) ) {
4850 $markerStart = strpos( $s, $this->mUniqPrefix
, $i );
4851 if ( $markerStart === false ) {
4852 $out .= call_user_func( $callback, substr( $s, $i ) );
4855 $out .= call_user_func( $callback, substr( $s, $i, $markerStart - $i ) );
4856 $markerEnd = strpos( $s, self
::MARKER_SUFFIX
, $markerStart );
4857 if ( $markerEnd === false ) {
4858 $out .= substr( $s, $markerStart );
4861 $markerEnd +
= strlen( self
::MARKER_SUFFIX
);
4862 $out .= substr( $s, $markerStart, $markerEnd - $markerStart );
4872 * @todo document, briefly.
4876 var $general, $nowiki;
4878 function __construct() {
4879 $this->general
= new ReplacementArray
;
4880 $this->nowiki
= new ReplacementArray
;
4883 function unstripGeneral( $text ) {
4884 wfProfileIn( __METHOD__
);
4887 $text = $this->general
->replace( $text );
4888 } while ( $text !== $oldText );
4889 wfProfileOut( __METHOD__
);
4893 function unstripNoWiki( $text ) {
4894 wfProfileIn( __METHOD__
);
4897 $text = $this->nowiki
->replace( $text );
4898 } while ( $text !== $oldText );
4899 wfProfileOut( __METHOD__
);
4903 function unstripBoth( $text ) {
4904 wfProfileIn( __METHOD__
);
4907 $text = $this->general
->replace( $text );
4908 $text = $this->nowiki
->replace( $text );
4909 } while ( $text !== $oldText );
4910 wfProfileOut( __METHOD__
);
4916 * @todo document, briefly.
4919 class OnlyIncludeReplacer
{
4922 function replace( $matches ) {
4923 if ( substr( $matches[1], -1 ) === "\n" ) {
4924 $this->output
.= substr( $matches[1], 0, -1 );
4926 $this->output
.= $matches[1];