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,
99 # Cleared with clearState():
100 var $mOutput, $mAutonumber, $mDTopen, $mStripState;
101 var $mIncludeCount, $mArgStack, $mLastSection, $mInPre;
102 var $mLinkHolders, $mLinkID;
103 var $mIncludeSizes, $mPPNodeCount, $mDefaultSort;
104 var $mTplExpandCache; // empty-frame expansion cache
105 var $mTplRedirCache, $mTplDomCache, $mHeadings, $mDoubleUnderscores;
106 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->mFunctionTagHooks
= array();
131 $this->mFunctionSynonyms
= array( 0 => array(), 1 => array() );
132 $this->mDefaultStripList
= $this->mStripList
= array( 'nowiki', 'gallery' );
133 $this->mUrlProtocols
= wfUrlProtocols();
134 $this->mExtLinkBracketedRegex
= '/\[(\b(' . wfUrlProtocols() . ')'.
135 '[^][<>"\\x00-\\x20\\x7F]+) *([^\]\\x0a\\x0d]*?)\]/S';
136 $this->mVarCache
= array();
137 if ( isset( $conf['preprocessorClass'] ) ) {
138 $this->mPreprocessorClass
= $conf['preprocessorClass'];
139 } elseif ( extension_loaded( 'domxml' ) ) {
140 // PECL extension that conflicts with the core DOM extension (bug 13770)
141 wfDebug( "Warning: you have the obsolete domxml extension for PHP. Please remove it!\n" );
142 $this->mPreprocessorClass
= 'Preprocessor_Hash';
143 } elseif ( extension_loaded( 'dom' ) ) {
144 $this->mPreprocessorClass
= 'Preprocessor_DOM';
146 $this->mPreprocessorClass
= 'Preprocessor_Hash';
148 $this->mMarkerIndex
= 0;
149 $this->mFirstCall
= true;
153 * Reduce memory usage to reduce the impact of circular references
155 function __destruct() {
156 if ( isset( $this->mLinkHolders
) ) {
157 $this->mLinkHolders
->__destruct();
159 foreach ( $this as $name => $value ) {
160 unset( $this->$name );
165 * Do various kinds of initialisation on the first call of the parser
167 function firstCallInit() {
168 if ( !$this->mFirstCall
) {
171 $this->mFirstCall
= false;
173 wfProfileIn( __METHOD__
);
175 $this->setHook( 'pre', array( $this, 'renderPreTag' ) );
176 CoreParserFunctions
::register( $this );
177 $this->initialiseVariables();
179 wfRunHooks( 'ParserFirstCallInit', array( &$this ) );
180 wfProfileOut( __METHOD__
);
188 function clearState() {
189 wfProfileIn( __METHOD__
);
190 if ( $this->mFirstCall
) {
191 $this->firstCallInit();
193 $this->mOutput
= new ParserOutput
;
194 $this->mAutonumber
= 0;
195 $this->mLastSection
= '';
196 $this->mDTopen
= false;
197 $this->mIncludeCount
= array();
198 $this->mStripState
= new StripState
;
199 $this->mArgStack
= false;
200 $this->mInPre
= false;
201 $this->mLinkHolders
= new LinkHolderArray( $this );
203 $this->mRevisionTimestamp
= $this->mRevisionId
= null;
206 * Prefix for temporary replacement strings for the multipass parser.
207 * \x07 should never appear in input as it's disallowed in XML.
208 * Using it at the front also gives us a little extra robustness
209 * since it shouldn't match when butted up against identifier-like
212 * Must not consist of all title characters, or else it will change
213 * the behaviour of <nowiki> in a link.
215 #$this->mUniqPrefix = "\x07UNIQ" . Parser::getRandomString();
216 # Changed to \x7f to allow XML double-parsing -- TS
217 $this->mUniqPrefix
= "\x7fUNIQ" . self
::getRandomString();
220 # Clear these on every parse, bug 4549
221 $this->mTplExpandCache
= $this->mTplRedirCache
= $this->mTplDomCache
= array();
223 $this->mShowToc
= true;
224 $this->mForceTocPosition
= false;
225 $this->mIncludeSizes
= array(
229 $this->mPPNodeCount
= 0;
230 $this->mDefaultSort
= false;
231 $this->mHeadings
= array();
232 $this->mDoubleUnderscores
= array();
233 $this->mExpensiveFunctionCount
= 0;
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 && $this->mOptions
->mTidy
) ||
$wgAlwaysUseTidy ) {
378 $text = MWTidy
::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 * If $frame is not provided, then template variables (e.g., {{{1}}}) within $text are not expanded
442 * @param $text String: text extension wants to have parsed
443 * @param PPFrame $frame: The frame to use for expanding any template variables
445 function recursiveTagParse( $text, $frame=false ) {
446 wfProfileIn( __METHOD__
);
447 wfRunHooks( 'ParserBeforeStrip', array( &$this, &$text, &$this->mStripState
) );
448 wfRunHooks( 'ParserAfterStrip', array( &$this, &$text, &$this->mStripState
) );
449 $text = $this->internalParse( $text, false, $frame );
450 wfProfileOut( __METHOD__
);
455 * Expand templates and variables in the text, producing valid, static wikitext.
456 * Also removes comments.
458 function preprocess( $text, $title, $options, $revid = null ) {
459 wfProfileIn( __METHOD__
);
461 $this->setOutputType( self
::OT_PREPROCESS
);
462 $this->mOptions
= $options;
463 $this->setTitle( $title );
464 if( $revid !== null ) {
465 $this->mRevisionId
= $revid;
467 wfRunHooks( 'ParserBeforeStrip', array( &$this, &$text, &$this->mStripState
) );
468 wfRunHooks( 'ParserAfterStrip', array( &$this, &$text, &$this->mStripState
) );
469 $text = $this->replaceVariables( $text );
470 $text = $this->mStripState
->unstripBoth( $text );
471 wfProfileOut( __METHOD__
);
476 * Get a random string
481 function getRandomString() {
482 return dechex(mt_rand(0, 0x7fffffff)) . dechex(mt_rand(0, 0x7fffffff));
485 function &getTitle() { return $this->mTitle
; }
486 function getOptions() { return $this->mOptions
; }
487 function getRevisionId() { return $this->mRevisionId
; }
488 function getOutput() { return $this->mOutput
; }
489 function nextLinkID() { return $this->mLinkID++
; }
491 function getFunctionLang() {
492 global $wgLang, $wgContLang;
494 $target = $this->mOptions
->getTargetLanguage();
495 if ( $target !== null ) {
498 return $this->mOptions
->getInterfaceMessage() ?
$wgLang : $wgContLang;
503 * Get a preprocessor object
505 function getPreprocessor() {
506 if ( !isset( $this->mPreprocessor
) ) {
507 $class = $this->mPreprocessorClass
;
508 $this->mPreprocessor
= new $class( $this );
510 return $this->mPreprocessor
;
514 * Replaces all occurrences of HTML-style comments and the given tags
515 * in the text with a random marker and returns the next text. The output
516 * parameter $matches will be an associative array filled with data in
518 * 'UNIQ-xxxxx' => array(
521 * array( 'param' => 'x' ),
522 * '<element param="x">tag content</element>' ) )
524 * @param $elements list of element names. Comments are always extracted.
525 * @param $text Source text string.
526 * @param $uniq_prefix
531 function extractTagsAndParams($elements, $text, &$matches, $uniq_prefix = ''){
536 $taglist = implode( '|', $elements );
537 $start = "/<($taglist)(\\s+[^>]*?|\\s*?)(\/?" . ">)|<(!--)/i";
539 while ( '' != $text ) {
540 $p = preg_split( $start, $text, 2, PREG_SPLIT_DELIM_CAPTURE
);
542 if( count( $p ) < 5 ) {
545 if( count( $p ) > 5 ) {
559 $marker = "$uniq_prefix-$element-" . sprintf('%08X', $n++
) . self
::MARKER_SUFFIX
;
560 $stripped .= $marker;
562 if ( $close === '/>' ) {
563 // Empty element tag, <tag />
568 if( $element === '!--' ) {
571 $end = "/(<\\/$element\\s*>)/i";
573 $q = preg_split( $end, $inside, 2, PREG_SPLIT_DELIM_CAPTURE
);
575 if( count( $q ) < 3 ) {
576 # No end tag -- let it run out to the end of the text.
585 $matches[$marker] = array( $element,
587 Sanitizer
::decodeTagAttributes( $attributes ),
588 "<$element$attributes$close$content$tail" );
594 * Get a list of strippable XML-like elements
596 function getStripList() {
598 $elements = $this->mStripList
;
600 $elements[] = 'html';
602 if( $this->mOptions
->getUseTeX() ) {
603 $elements[] = 'math';
609 * @deprecated use replaceVariables
611 function strip( $text, $state, $stripcomments = false , $dontstrip = array () ) {
616 * Restores pre, math, and other extensions removed by strip()
618 * always call unstripNoWiki() after this one
620 * @deprecated use $this->mStripState->unstrip()
622 function unstrip( $text, $state ) {
623 return $state->unstripGeneral( $text );
627 * Always call this after unstrip() to preserve the order
630 * @deprecated use $this->mStripState->unstrip()
632 function unstripNoWiki( $text, $state ) {
633 return $state->unstripNoWiki( $text );
637 * @deprecated use $this->mStripState->unstripBoth()
639 function unstripForHTML( $text ) {
640 return $this->mStripState
->unstripBoth( $text );
644 * Add an item to the strip state
645 * Returns the unique tag which must be inserted into the stripped text
646 * The tag will be replaced with the original text in unstrip()
650 function insertStripItem( $text ) {
651 $rnd = "{$this->mUniqPrefix}-item-{$this->mMarkerIndex}-" . self
::MARKER_SUFFIX
;
652 $this->mMarkerIndex++
;
653 $this->mStripState
->general
->setPair( $rnd, $text );
658 * Interface with html tidy
659 * @deprecated Use MWTidy::tidy()
661 public static function tidy( $text ) {
662 wfDeprecated( __METHOD__
);
663 return MWTidy
::tidy( $text );
667 * parse the wiki syntax used to render tables
671 function doTableStuff ( $text ) {
672 wfProfileIn( __METHOD__
);
674 $lines = StringUtils
::explode( "\n", $text );
676 $td_history = array (); // Is currently a td tag open?
677 $last_tag_history = array (); // Save history of last lag activated (td, th or caption)
678 $tr_history = array (); // Is currently a tr tag open?
679 $tr_attributes = array (); // history of tr attributes
680 $has_opened_tr = array(); // Did this table open a <tr> element?
681 $indent_level = 0; // indent level of the table
683 foreach ( $lines as $outLine ) {
684 $line = trim( $outLine );
686 if( $line == '' ) { // empty line, go to next line
687 $out .= $outLine."\n";
690 $first_character = $line[0];
693 if ( preg_match( '/^(:*)\{\|(.*)$/', $line , $matches ) ) {
694 // First check if we are starting a new table
695 $indent_level = strlen( $matches[1] );
697 $attributes = $this->mStripState
->unstripBoth( $matches[2] );
698 $attributes = Sanitizer
::fixTagAttributes ( $attributes , 'table' );
700 $outLine = str_repeat( '<dl><dd>' , $indent_level ) . "<table{$attributes}>";
701 array_push ( $td_history , false );
702 array_push ( $last_tag_history , '' );
703 array_push ( $tr_history , false );
704 array_push ( $tr_attributes , '' );
705 array_push ( $has_opened_tr , false );
706 } else if ( count ( $td_history ) == 0 ) {
707 // Don't do any of the following
708 $out .= $outLine."\n";
710 } else if ( substr ( $line , 0 , 2 ) === '|}' ) {
711 // We are ending a table
712 $line = '</table>' . substr ( $line , 2 );
713 $last_tag = array_pop ( $last_tag_history );
715 if ( !array_pop ( $has_opened_tr ) ) {
716 $line = "<tr><td></td></tr>{$line}";
719 if ( array_pop ( $tr_history ) ) {
720 $line = "</tr>{$line}";
723 if ( array_pop ( $td_history ) ) {
724 $line = "</{$last_tag}>{$line}";
726 array_pop ( $tr_attributes );
727 $outLine = $line . str_repeat( '</dd></dl>' , $indent_level );
728 } else if ( substr ( $line , 0 , 2 ) === '|-' ) {
729 // Now we have a table row
730 $line = preg_replace( '#^\|-+#', '', $line );
732 // Whats after the tag is now only attributes
733 $attributes = $this->mStripState
->unstripBoth( $line );
734 $attributes = Sanitizer
::fixTagAttributes ( $attributes , 'tr' );
735 array_pop ( $tr_attributes );
736 array_push ( $tr_attributes , $attributes );
739 $last_tag = array_pop ( $last_tag_history );
740 array_pop ( $has_opened_tr );
741 array_push ( $has_opened_tr , true );
743 if ( array_pop ( $tr_history ) ) {
747 if ( array_pop ( $td_history ) ) {
748 $line = "</{$last_tag}>{$line}";
752 array_push ( $tr_history , false );
753 array_push ( $td_history , false );
754 array_push ( $last_tag_history , '' );
756 else if ( $first_character === '|' ||
$first_character === '!' ||
substr ( $line , 0 , 2 ) === '|+' ) {
757 // This might be cell elements, td, th or captions
758 if ( substr ( $line , 0 , 2 ) === '|+' ) {
759 $first_character = '+';
760 $line = substr ( $line , 1 );
763 $line = substr ( $line , 1 );
765 if ( $first_character === '!' ) {
766 $line = str_replace ( '!!' , '||' , $line );
769 // Split up multiple cells on the same line.
770 // FIXME : This can result in improper nesting of tags processed
771 // by earlier parser steps, but should avoid splitting up eg
772 // attribute values containing literal "||".
773 $cells = StringUtils
::explodeMarkup( '||' , $line );
777 // Loop through each table cell
778 foreach ( $cells as $cell )
781 if ( $first_character !== '+' )
783 $tr_after = array_pop ( $tr_attributes );
784 if ( !array_pop ( $tr_history ) ) {
785 $previous = "<tr{$tr_after}>\n";
787 array_push ( $tr_history , true );
788 array_push ( $tr_attributes , '' );
789 array_pop ( $has_opened_tr );
790 array_push ( $has_opened_tr , true );
793 $last_tag = array_pop ( $last_tag_history );
795 if ( array_pop ( $td_history ) ) {
796 $previous = "</{$last_tag}>{$previous}";
799 if ( $first_character === '|' ) {
801 } else if ( $first_character === '!' ) {
803 } else if ( $first_character === '+' ) {
804 $last_tag = 'caption';
809 array_push ( $last_tag_history , $last_tag );
811 // A cell could contain both parameters and data
812 $cell_data = explode ( '|' , $cell , 2 );
814 // Bug 553: Note that a '|' inside an invalid link should not
815 // be mistaken as delimiting cell parameters
816 if ( strpos( $cell_data[0], '[[' ) !== false ) {
817 $cell = "{$previous}<{$last_tag}>{$cell}";
818 } else if ( count ( $cell_data ) == 1 )
819 $cell = "{$previous}<{$last_tag}>{$cell_data[0]}";
821 $attributes = $this->mStripState
->unstripBoth( $cell_data[0] );
822 $attributes = Sanitizer
::fixTagAttributes( $attributes , $last_tag );
823 $cell = "{$previous}<{$last_tag}{$attributes}>{$cell_data[1]}";
827 array_push ( $td_history , true );
830 $out .= $outLine . "\n";
833 // Closing open td, tr && table
834 while ( count ( $td_history ) > 0 )
836 if ( array_pop ( $td_history ) ) {
839 if ( array_pop ( $tr_history ) ) {
842 if ( !array_pop ( $has_opened_tr ) ) {
843 $out .= "<tr><td></td></tr>\n" ;
846 $out .= "</table>\n";
849 // Remove trailing line-ending (b/c)
850 if ( substr( $out, -1 ) === "\n" ) {
851 $out = substr( $out, 0, -1 );
854 // special case: don't return empty table
855 if( $out === "<table>\n<tr><td></td></tr>\n</table>" ) {
859 wfProfileOut( __METHOD__
);
865 * Helper function for parse() that transforms wiki markup into
866 * HTML. Only called for $mOutputType == self::OT_HTML.
870 function internalParse( $text, $isMain = true, $frame=false ) {
871 wfProfileIn( __METHOD__
);
875 # Hook to suspend the parser in this state
876 if ( !wfRunHooks( 'ParserBeforeInternalParse', array( &$this, &$text, &$this->mStripState
) ) ) {
877 wfProfileOut( __METHOD__
);
881 // if $frame is provided, then use $frame for replacing any variables
883 // use frame depth to infer how include/noinclude tags should be handled
884 // depth=0 means this is the top-level document; otherwise it's an included document
888 $flag = Parser
::PTD_FOR_INCLUSION
;
889 $dom = $this->preprocessToDom( $text, $flag );
890 $text = $frame->expand( $dom );
892 // if $frame is not provided, then use old-style replaceVariables
894 $text = $this->replaceVariables( $text );
897 $text = Sanitizer
::removeHTMLtags( $text, array( &$this, 'attributeStripCallback' ), false, array_keys( $this->mTransparentTagHooks
) );
898 wfRunHooks( 'InternalParseBeforeLinks', array( &$this, &$text, &$this->mStripState
) );
900 // Tables need to come after variable replacement for things to work
901 // properly; putting them before other transformations should keep
902 // exciting things like link expansions from showing up in surprising
904 $text = $this->doTableStuff( $text );
906 $text = preg_replace( '/(^|\n)-----*/', '\\1<hr />', $text );
908 $text = $this->doDoubleUnderscore( $text );
909 $text = $this->doHeadings( $text );
910 if( $this->mOptions
->getUseDynamicDates() ) {
911 $df = DateFormatter
::getInstance();
912 $text = $df->reformat( $this->mOptions
->getDateFormat(), $text );
914 $text = $this->doAllQuotes( $text );
915 $text = $this->replaceInternalLinks( $text );
916 $text = $this->replaceExternalLinks( $text );
918 # replaceInternalLinks may sometimes leave behind
919 # absolute URLs, which have to be masked to hide them from replaceExternalLinks
920 $text = str_replace($this->mUniqPrefix
.'NOPARSE', '', $text);
922 $text = $this->doMagicLinks( $text );
923 $text = $this->formatHeadings( $text, $origText, $isMain );
925 wfProfileOut( __METHOD__
);
930 * Replace special strings like "ISBN xxx" and "RFC xxx" with
931 * magic external links.
936 function doMagicLinks( $text ) {
937 wfProfileIn( __METHOD__
);
938 $prots = $this->mUrlProtocols
;
939 $urlChar = self
::EXT_LINK_URL_CLASS
;
940 $text = preg_replace_callback(
942 (<a.*?</a>) | # m[1]: Skip link text
943 (<.*?>) | # m[2]: Skip stuff inside HTML elements' . "
944 (\\b(?:$prots)$urlChar+) | # m[3]: Free external links" . '
945 (?:RFC|PMID)\s+([0-9]+) | # m[4]: RFC or PMID, capture number
946 ISBN\s+(\b # m[5]: ISBN, capture number
947 (?: 97[89] [\ \-]? )? # optional 13-digit ISBN prefix
948 (?: [0-9] [\ \-]? ){9} # 9 digits with opt. delimiters
949 [0-9Xx] # check digit
951 )!x', array( &$this, 'magicLinkCallback' ), $text );
952 wfProfileOut( __METHOD__
);
956 function magicLinkCallback( $m ) {
957 if ( isset( $m[1] ) && $m[1] !== '' ) {
960 } elseif ( isset( $m[2] ) && $m[2] !== '' ) {
963 } elseif ( isset( $m[3] ) && $m[3] !== '' ) {
965 return $this->makeFreeExternalLink( $m[0] );
966 } elseif ( isset( $m[4] ) && $m[4] !== '' ) {
969 if ( substr( $m[0], 0, 3 ) === 'RFC' ) {
972 $CssClass = 'mw-magiclink-rfc';
974 } elseif ( substr( $m[0], 0, 4 ) === 'PMID' ) {
976 $urlmsg = 'pubmedurl';
977 $CssClass = 'mw-magiclink-pmid';
980 throw new MWException( __METHOD__
.': unrecognised match type "' .
981 substr($m[0], 0, 20 ) . '"' );
983 $url = wfMsg( $urlmsg, $id);
984 $sk = $this->mOptions
->getSkin();
985 $la = $sk->getExternalLinkAttributes( "external $CssClass" );
986 return "<a href=\"{$url}\"{$la}>{$keyword} {$id}</a>";
987 } elseif ( isset( $m[5] ) && $m[5] !== '' ) {
990 $num = strtr( $isbn, array(
995 $titleObj = SpecialPage
::getTitleFor( 'Booksources', $num );
997 $titleObj->escapeLocalUrl() .
998 "\" class=\"internal mw-magiclink-isbn\">ISBN $isbn</a>";
1005 * Make a free external link, given a user-supplied URL
1009 function makeFreeExternalLink( $url ) {
1011 wfProfileIn( __METHOD__
);
1013 $sk = $this->mOptions
->getSkin();
1016 # The characters '<' and '>' (which were escaped by
1017 # removeHTMLtags()) should not be included in
1018 # URLs, per RFC 2396.
1020 if (preg_match('/&(lt|gt);/', $url, $m2, PREG_OFFSET_CAPTURE
)) {
1021 $trail = substr($url, $m2[0][1]) . $trail;
1022 $url = substr($url, 0, $m2[0][1]);
1025 # Move trailing punctuation to $trail
1027 # If there is no left bracket, then consider right brackets fair game too
1028 if ( strpos( $url, '(' ) === false ) {
1032 $numSepChars = strspn( strrev( $url ), $sep );
1033 if ( $numSepChars ) {
1034 $trail = substr( $url, -$numSepChars ) . $trail;
1035 $url = substr( $url, 0, -$numSepChars );
1038 $url = Sanitizer
::cleanUrl( $url );
1040 # Is this an external image?
1041 $text = $this->maybeMakeExternalImage( $url );
1042 if ( $text === false ) {
1043 # Not an image, make a link
1044 $text = $sk->makeExternalLink( $url, $wgContLang->markNoConversion($url), true, 'free',
1045 $this->getExternalLinkAttribs( $url ) );
1046 # Register it in the output object...
1047 # Replace unnecessary URL escape codes with their equivalent characters
1048 $pasteurized = self
::replaceUnusualEscapes( $url );
1049 $this->mOutput
->addExternalLink( $pasteurized );
1051 wfProfileOut( __METHOD__
);
1052 return $text . $trail;
1057 * Parse headers and return html
1061 function doHeadings( $text ) {
1062 wfProfileIn( __METHOD__
);
1063 for ( $i = 6; $i >= 1; --$i ) {
1064 $h = str_repeat( '=', $i );
1065 $text = preg_replace( "/^$h(.+)$h\\s*$/m",
1066 "<h$i>\\1</h$i>", $text );
1068 wfProfileOut( __METHOD__
);
1073 * Replace single quotes with HTML markup
1075 * @return string the altered text
1077 function doAllQuotes( $text ) {
1078 wfProfileIn( __METHOD__
);
1080 $lines = StringUtils
::explode( "\n", $text );
1081 foreach ( $lines as $line ) {
1082 $outtext .= $this->doQuotes( $line ) . "\n";
1084 $outtext = substr($outtext, 0,-1);
1085 wfProfileOut( __METHOD__
);
1090 * Helper function for doAllQuotes()
1092 public function doQuotes( $text ) {
1093 $arr = preg_split( "/(''+)/", $text, -1, PREG_SPLIT_DELIM_CAPTURE
);
1094 if ( count( $arr ) == 1 )
1098 # First, do some preliminary work. This may shift some apostrophes from
1099 # being mark-up to being text. It also counts the number of occurrences
1100 # of bold and italics mark-ups.
1104 foreach ( $arr as $r )
1106 if ( ( $i %
2 ) == 1 )
1108 # If there are ever four apostrophes, assume the first is supposed to
1109 # be text, and the remaining three constitute mark-up for bold text.
1110 if ( strlen( $arr[$i] ) == 4 )
1115 # If there are more than 5 apostrophes in a row, assume they're all
1116 # text except for the last 5.
1117 else if ( strlen( $arr[$i] ) > 5 )
1119 $arr[$i-1] .= str_repeat( "'", strlen( $arr[$i] ) - 5 );
1122 # Count the number of occurrences of bold and italics mark-ups.
1123 # We are not counting sequences of five apostrophes.
1124 if ( strlen( $arr[$i] ) == 2 ) { $numitalics++
; }
1125 else if ( strlen( $arr[$i] ) == 3 ) { $numbold++
; }
1126 else if ( strlen( $arr[$i] ) == 5 ) { $numitalics++
; $numbold++
; }
1131 # If there is an odd number of both bold and italics, it is likely
1132 # that one of the bold ones was meant to be an apostrophe followed
1133 # by italics. Which one we cannot know for certain, but it is more
1134 # likely to be one that has a single-letter word before it.
1135 if ( ( $numbold %
2 == 1 ) && ( $numitalics %
2 == 1 ) )
1138 $firstsingleletterword = -1;
1139 $firstmultiletterword = -1;
1141 foreach ( $arr as $r )
1143 if ( ( $i %
2 == 1 ) and ( strlen( $r ) == 3 ) )
1145 $x1 = substr ($arr[$i-1], -1);
1146 $x2 = substr ($arr[$i-1], -2, 1);
1148 if ($firstspace == -1) $firstspace = $i;
1149 } else if ($x2 === ' ') {
1150 if ($firstsingleletterword == -1) $firstsingleletterword = $i;
1152 if ($firstmultiletterword == -1) $firstmultiletterword = $i;
1158 # If there is a single-letter word, use it!
1159 if ($firstsingleletterword > -1)
1161 $arr [ $firstsingleletterword ] = "''";
1162 $arr [ $firstsingleletterword-1 ] .= "'";
1164 # If not, but there's a multi-letter word, use that one.
1165 else if ($firstmultiletterword > -1)
1167 $arr [ $firstmultiletterword ] = "''";
1168 $arr [ $firstmultiletterword-1 ] .= "'";
1170 # ... otherwise use the first one that has neither.
1171 # (notice that it is possible for all three to be -1 if, for example,
1172 # there is only one pentuple-apostrophe in the line)
1173 else if ($firstspace > -1)
1175 $arr [ $firstspace ] = "''";
1176 $arr [ $firstspace-1 ] .= "'";
1180 # Now let's actually convert our apostrophic mush to HTML!
1185 foreach ($arr as $r)
1189 if ($state === 'both')
1196 if (strlen ($r) == 2)
1199 { $output .= '</i>'; $state = ''; }
1200 else if ($state === 'bi')
1201 { $output .= '</i>'; $state = 'b'; }
1202 else if ($state === 'ib')
1203 { $output .= '</b></i><b>'; $state = 'b'; }
1204 else if ($state === 'both')
1205 { $output .= '<b><i>'.$buffer.'</i>'; $state = 'b'; }
1206 else # $state can be 'b' or ''
1207 { $output .= '<i>'; $state .= 'i'; }
1209 else if (strlen ($r) == 3)
1212 { $output .= '</b>'; $state = ''; }
1213 else if ($state === 'bi')
1214 { $output .= '</i></b><i>'; $state = 'i'; }
1215 else if ($state === 'ib')
1216 { $output .= '</b>'; $state = 'i'; }
1217 else if ($state === 'both')
1218 { $output .= '<i><b>'.$buffer.'</b>'; $state = 'i'; }
1219 else # $state can be 'i' or ''
1220 { $output .= '<b>'; $state .= 'b'; }
1222 else if (strlen ($r) == 5)
1225 { $output .= '</b><i>'; $state = 'i'; }
1226 else if ($state === 'i')
1227 { $output .= '</i><b>'; $state = 'b'; }
1228 else if ($state === 'bi')
1229 { $output .= '</i></b>'; $state = ''; }
1230 else if ($state === 'ib')
1231 { $output .= '</b></i>'; $state = ''; }
1232 else if ($state === 'both')
1233 { $output .= '<i><b>'.$buffer.'</b></i>'; $state = ''; }
1234 else # ($state == '')
1235 { $buffer = ''; $state = 'both'; }
1240 # Now close all remaining tags. Notice that the order is important.
1241 if ($state === 'b' ||
$state === 'ib')
1243 if ($state === 'i' ||
$state === 'bi' ||
$state === 'ib')
1245 if ($state === 'bi')
1247 # There might be lonely ''''', so make sure we have a buffer
1248 if ($state === 'both' && $buffer)
1249 $output .= '<b><i>'.$buffer.'</i></b>';
1255 * Replace external links (REL)
1257 * Note: this is all very hackish and the order of execution matters a lot.
1258 * Make sure to run maintenance/parserTests.php if you change this code.
1262 function replaceExternalLinks( $text ) {
1264 wfProfileIn( __METHOD__
);
1266 $sk = $this->mOptions
->getSkin();
1268 $bits = preg_split( $this->mExtLinkBracketedRegex
, $text, -1, PREG_SPLIT_DELIM_CAPTURE
);
1269 $s = array_shift( $bits );
1272 while ( $i<count( $bits ) ) {
1274 $protocol = $bits[$i++
];
1275 $text = $bits[$i++
];
1276 $trail = $bits[$i++
];
1278 # The characters '<' and '>' (which were escaped by
1279 # removeHTMLtags()) should not be included in
1280 # URLs, per RFC 2396.
1282 if (preg_match('/&(lt|gt);/', $url, $m2, PREG_OFFSET_CAPTURE
)) {
1283 $text = substr($url, $m2[0][1]) . ' ' . $text;
1284 $url = substr($url, 0, $m2[0][1]);
1287 # If the link text is an image URL, replace it with an <img> tag
1288 # This happened by accident in the original parser, but some people used it extensively
1289 $img = $this->maybeMakeExternalImage( $text );
1290 if ( $img !== false ) {
1296 # Set linktype for CSS - if URL==text, link is essentially free
1297 $linktype = ($text === $url) ?
'free' : 'text';
1299 # No link text, e.g. [http://domain.tld/some.link]
1300 if ( $text == '' ) {
1301 # Autonumber if allowed. See bug #5918
1302 if ( strpos( wfUrlProtocols(), substr($protocol, 0, strpos($protocol, ':')) ) !== false ) {
1303 $langObj = $this->getFunctionLang();
1304 $text = '[' . $langObj->formatNum( ++
$this->mAutonumber
) . ']';
1305 $linktype = 'autonumber';
1307 # Otherwise just use the URL
1308 $text = htmlspecialchars( $url );
1312 # Have link text, e.g. [http://domain.tld/some.link text]s
1314 list( $dtrail, $trail ) = Linker
::splitTrail( $trail );
1317 $text = $wgContLang->markNoConversion($text);
1319 $url = Sanitizer
::cleanUrl( $url );
1321 # Use the encoded URL
1322 # This means that users can paste URLs directly into the text
1323 # Funny characters like ö aren't valid in URLs anyway
1324 # This was changed in August 2004
1325 $s .= $sk->makeExternalLink( $url, $text, false, $linktype,
1326 $this->getExternalLinkAttribs( $url ) ) . $dtrail . $trail;
1328 # Register link in the output object.
1329 # Replace unnecessary URL escape codes with the referenced character
1330 # This prevents spammers from hiding links from the filters
1331 $pasteurized = self
::replaceUnusualEscapes( $url );
1332 $this->mOutput
->addExternalLink( $pasteurized );
1335 wfProfileOut( __METHOD__
);
1340 * Get an associative array of additional HTML attributes appropriate for a
1341 * particular external link. This currently may include rel => nofollow
1342 * (depending on configuration, namespace, and the URL's domain) and/or a
1343 * target attribute (depending on configuration).
1345 * @param string $url Optional URL, to extract the domain from for rel =>
1346 * nofollow if appropriate
1347 * @return array Associative array of HTML attributes
1349 function getExternalLinkAttribs( $url = false ) {
1351 global $wgNoFollowLinks, $wgNoFollowNsExceptions;
1352 $ns = $this->mTitle
->getNamespace();
1353 if( $wgNoFollowLinks && !in_array($ns, $wgNoFollowNsExceptions) ) {
1354 $attribs['rel'] = 'nofollow';
1356 global $wgNoFollowDomainExceptions;
1357 if ( $wgNoFollowDomainExceptions ) {
1358 $bits = wfParseUrl( $url );
1359 if ( is_array( $bits ) && isset( $bits['host'] ) ) {
1360 foreach ( $wgNoFollowDomainExceptions as $domain ) {
1361 if( substr( $bits['host'], -strlen( $domain ) )
1363 unset( $attribs['rel'] );
1370 if ( $this->mOptions
->getExternalLinkTarget() ) {
1371 $attribs['target'] = $this->mOptions
->getExternalLinkTarget();
1378 * Replace unusual URL escape codes with their equivalent characters
1382 * @todo This can merge genuinely required bits in the path or query string,
1383 * breaking legit URLs. A proper fix would treat the various parts of
1384 * the URL differently; as a workaround, just use the output for
1385 * statistical records, not for actual linking/output.
1387 static function replaceUnusualEscapes( $url ) {
1388 return preg_replace_callback( '/%[0-9A-Fa-f]{2}/',
1389 array( __CLASS__
, 'replaceUnusualEscapesCallback' ), $url );
1393 * Callback function used in replaceUnusualEscapes().
1394 * Replaces unusual URL escape codes with their equivalent character
1398 private static function replaceUnusualEscapesCallback( $matches ) {
1399 $char = urldecode( $matches[0] );
1400 $ord = ord( $char );
1401 // Is it an unsafe or HTTP reserved character according to RFC 1738?
1402 if ( $ord > 32 && $ord < 127 && strpos( '<>"#{}|\^~[]`;/?', $char ) === false ) {
1403 // No, shouldn't be escaped
1406 // Yes, leave it escaped
1412 * make an image if it's allowed, either through the global
1413 * option, through the exception, or through the on-wiki whitelist
1416 function maybeMakeExternalImage( $url ) {
1417 $sk = $this->mOptions
->getSkin();
1418 $imagesfrom = $this->mOptions
->getAllowExternalImagesFrom();
1419 $imagesexception = !empty($imagesfrom);
1421 # $imagesfrom could be either a single string or an array of strings, parse out the latter
1422 if( $imagesexception && is_array( $imagesfrom ) ) {
1423 $imagematch = false;
1424 foreach( $imagesfrom as $match ) {
1425 if( strpos( $url, $match ) === 0 ) {
1430 } elseif( $imagesexception ) {
1431 $imagematch = (strpos( $url, $imagesfrom ) === 0);
1433 $imagematch = false;
1435 if ( $this->mOptions
->getAllowExternalImages()
1436 ||
( $imagesexception && $imagematch ) ) {
1437 if ( preg_match( self
::EXT_IMAGE_REGEX
, $url ) ) {
1439 $text = $sk->makeExternalImage( $url );
1442 if( !$text && $this->mOptions
->getEnableImageWhitelist()
1443 && preg_match( self
::EXT_IMAGE_REGEX
, $url ) ) {
1444 $whitelist = explode( "\n", wfMsgForContent( 'external_image_whitelist' ) );
1445 foreach( $whitelist as $entry ) {
1446 # Sanitize the regex fragment, make it case-insensitive, ignore blank entries/comments
1447 if( strpos( $entry, '#' ) === 0 ||
$entry === '' )
1449 if( preg_match( '/' . str_replace( '/', '\\/', $entry ) . '/i', $url ) ) {
1450 # Image matches a whitelist entry
1451 $text = $sk->makeExternalImage( $url );
1460 * Process [[ ]] wikilinks
1461 * @return processed text
1465 function replaceInternalLinks( $s ) {
1466 $this->mLinkHolders
->merge( $this->replaceInternalLinks2( $s ) );
1471 * Process [[ ]] wikilinks (RIL)
1472 * @return LinkHolderArray
1476 function replaceInternalLinks2( &$s ) {
1479 wfProfileIn( __METHOD__
);
1481 wfProfileIn( __METHOD__
.'-setup' );
1482 static $tc = FALSE, $e1, $e1_img;
1483 # the % is needed to support urlencoded titles as well
1485 $tc = Title
::legalChars() . '#%';
1486 # Match a link having the form [[namespace:link|alternate]]trail
1487 $e1 = "/^([{$tc}]+)(?:\\|(.+?))?]](.*)\$/sD";
1488 # Match cases where there is no "]]", which might still be images
1489 $e1_img = "/^([{$tc}]+)\\|(.*)\$/sD";
1492 $sk = $this->mOptions
->getSkin();
1493 $holders = new LinkHolderArray( $this );
1495 #split the entire text string on occurences of [[
1496 $a = StringUtils
::explode( '[[', ' ' . $s );
1497 #get the first element (all text up to first [[), and remove the space we added
1500 $line = $a->current(); # Workaround for broken ArrayIterator::next() that returns "void"
1501 $s = substr( $s, 1 );
1503 $useLinkPrefixExtension = $wgContLang->linkPrefixExtension();
1505 if ( $useLinkPrefixExtension ) {
1506 # Match the end of a line for a word that's not followed by whitespace,
1507 # e.g. in the case of 'The Arab al[[Razi]]', 'al' will be matched
1508 $e2 = wfMsgForContent( 'linkprefix' );
1511 if( is_null( $this->mTitle
) ) {
1512 wfProfileOut( __METHOD__
.'-setup' );
1513 wfProfileOut( __METHOD__
);
1514 throw new MWException( __METHOD__
.": \$this->mTitle is null\n" );
1516 $nottalk = !$this->mTitle
->isTalkPage();
1518 if ( $useLinkPrefixExtension ) {
1520 if ( preg_match( $e2, $s, $m ) ) {
1521 $first_prefix = $m[2];
1523 $first_prefix = false;
1529 if($wgContLang->hasVariants()) {
1530 $selflink = $wgContLang->convertLinkToAllVariants($this->mTitle
->getPrefixedText());
1532 $selflink = array($this->mTitle
->getPrefixedText());
1534 $useSubpages = $this->areSubpagesAllowed();
1535 wfProfileOut( __METHOD__
.'-setup' );
1537 # Loop for each link
1538 for ( ; $line !== false && $line !== null ; $a->next(), $line = $a->current() ) {
1539 # Check for excessive memory usage
1540 if ( $holders->isBig() ) {
1542 # Do the existence check, replace the link holders and clear the array
1543 $holders->replace( $s );
1547 if ( $useLinkPrefixExtension ) {
1548 wfProfileIn( __METHOD__
.'-prefixhandling' );
1549 if ( preg_match( $e2, $s, $m ) ) {
1557 $prefix = $first_prefix;
1558 $first_prefix = false;
1560 wfProfileOut( __METHOD__
.'-prefixhandling' );
1563 $might_be_img = false;
1565 wfProfileIn( __METHOD__
."-e1" );
1566 if ( preg_match( $e1, $line, $m ) ) { # page with normal text or alt
1568 # If we get a ] at the beginning of $m[3] that means we have a link that's something like:
1569 # [[Image:Foo.jpg|[http://example.com desc]]] <- having three ] in a row fucks up,
1570 # the real problem is with the $e1 regex
1573 # Still some problems for cases where the ] is meant to be outside punctuation,
1574 # and no image is in sight. See bug 2095.
1577 substr( $m[3], 0, 1 ) === ']' &&
1578 strpos($text, '[') !== false
1581 $text .= ']'; # so that replaceExternalLinks($text) works later
1582 $m[3] = substr( $m[3], 1 );
1584 # fix up urlencoded title texts
1585 if( strpos( $m[1], '%' ) !== false ) {
1586 # Should anchors '#' also be rejected?
1587 $m[1] = str_replace( array('<', '>'), array('<', '>'), urldecode($m[1]) );
1590 } elseif( preg_match($e1_img, $line, $m) ) { # Invalid, but might be an image with a link in its caption
1591 $might_be_img = true;
1593 if ( strpos( $m[1], '%' ) !== false ) {
1594 $m[1] = urldecode($m[1]);
1597 } else { # Invalid form; output directly
1598 $s .= $prefix . '[[' . $line ;
1599 wfProfileOut( __METHOD__
."-e1" );
1602 wfProfileOut( __METHOD__
."-e1" );
1603 wfProfileIn( __METHOD__
."-misc" );
1605 # Don't allow internal links to pages containing
1606 # PROTO: where PROTO is a valid URL protocol; these
1607 # should be external links.
1608 if ( preg_match( '/^\b(?:' . wfUrlProtocols() . ')/', $m[1] ) ) {
1609 $s .= $prefix . '[[' . $line ;
1610 wfProfileOut( __METHOD__
."-misc" );
1614 # Make subpage if necessary
1615 if ( $useSubpages ) {
1616 $link = $this->maybeDoSubpageLink( $m[1], $text );
1621 $noforce = (substr( $m[1], 0, 1 ) !== ':');
1623 # Strip off leading ':'
1624 $link = substr( $link, 1 );
1627 wfProfileOut( __METHOD__
."-misc" );
1628 wfProfileIn( __METHOD__
."-title" );
1629 $nt = Title
::newFromText( $this->mStripState
->unstripNoWiki( $link ) );
1630 if ( $nt === NULL ) {
1631 $s .= $prefix . '[[' . $line;
1632 wfProfileOut( __METHOD__
."-title" );
1636 $ns = $nt->getNamespace();
1637 $iw = $nt->getInterWiki();
1638 wfProfileOut( __METHOD__
."-title" );
1640 if ( $might_be_img ) { # if this is actually an invalid link
1641 wfProfileIn( __METHOD__
."-might_be_img" );
1642 if ( $ns == NS_FILE
&& $noforce ) { #but might be an image
1645 #look at the next 'line' to see if we can close it there
1647 $next_line = $a->current();
1648 if ( $next_line === false ||
$next_line === null ) {
1651 $m = explode( ']]', $next_line, 3 );
1652 if ( count( $m ) == 3 ) {
1653 # the first ]] closes the inner link, the second the image
1655 $text .= "[[{$m[0]}]]{$m[1]}";
1658 } elseif ( count( $m ) == 2 ) {
1659 #if there's exactly one ]] that's fine, we'll keep looking
1660 $text .= "[[{$m[0]}]]{$m[1]}";
1662 #if $next_line is invalid too, we need look no further
1663 $text .= '[[' . $next_line;
1668 # we couldn't find the end of this imageLink, so output it raw
1669 #but don't ignore what might be perfectly normal links in the text we've examined
1670 $holders->merge( $this->replaceInternalLinks2( $text ) );
1671 $s .= "{$prefix}[[$link|$text";
1672 # note: no $trail, because without an end, there *is* no trail
1673 wfProfileOut( __METHOD__
."-might_be_img" );
1676 } else { #it's not an image, so output it raw
1677 $s .= "{$prefix}[[$link|$text";
1678 # note: no $trail, because without an end, there *is* no trail
1679 wfProfileOut( __METHOD__
."-might_be_img" );
1682 wfProfileOut( __METHOD__
."-might_be_img" );
1685 $wasblank = ( '' == $text );
1686 if ( $wasblank ) $text = $link;
1688 # Link not escaped by : , create the various objects
1692 wfProfileIn( __METHOD__
."-interwiki" );
1693 if ( $iw && $this->mOptions
->getInterwikiMagic() && $nottalk && $wgContLang->getLanguageName( $iw ) ) {
1694 $this->mOutput
->addLanguageLink( $nt->getFullText() );
1695 $s = rtrim($s . $prefix);
1696 $s .= trim($trail, "\n") == '' ?
'': $prefix . $trail;
1697 wfProfileOut( __METHOD__
."-interwiki" );
1700 wfProfileOut( __METHOD__
."-interwiki" );
1702 if ( $ns == NS_FILE
) {
1703 wfProfileIn( __METHOD__
."-image" );
1704 if ( !wfIsBadImage( $nt->getDBkey(), $this->mTitle
) ) {
1706 # if no parameters were passed, $text
1707 # becomes something like "File:Foo.png",
1708 # which we don't want to pass on to the
1712 # recursively parse links inside the image caption
1713 # actually, this will parse them in any other parameters, too,
1714 # but it might be hard to fix that, and it doesn't matter ATM
1715 $text = $this->replaceExternalLinks($text);
1716 $holders->merge( $this->replaceInternalLinks2( $text ) );
1718 # cloak any absolute URLs inside the image markup, so replaceExternalLinks() won't touch them
1719 $s .= $prefix . $this->armorLinks( $this->makeImage( $nt, $text, $holders ) ) . $trail;
1721 $this->mOutput
->addImage( $nt->getDBkey() );
1722 wfProfileOut( __METHOD__
."-image" );
1727 if ( $ns == NS_CATEGORY
) {
1728 wfProfileIn( __METHOD__
."-category" );
1729 $s = rtrim($s . "\n"); # bug 87
1732 $sortkey = $this->getDefaultSort();
1736 $sortkey = Sanitizer
::decodeCharReferences( $sortkey );
1737 $sortkey = str_replace( "\n", '', $sortkey );
1738 $sortkey = $wgContLang->convertCategoryKey( $sortkey );
1739 $this->mOutput
->addCategory( $nt->getDBkey(), $sortkey );
1742 * Strip the whitespace Category links produce, see bug 87
1743 * @todo We might want to use trim($tmp, "\n") here.
1745 $s .= trim($prefix . $trail, "\n") == '' ?
'': $prefix . $trail;
1747 wfProfileOut( __METHOD__
."-category" );
1752 # Self-link checking
1753 if( $nt->getFragment() === '' && $ns != NS_SPECIAL
) {
1754 if( in_array( $nt->getPrefixedText(), $selflink, true ) ) {
1755 $s .= $prefix . $sk->makeSelfLinkObj( $nt, $text, '', $trail );
1760 # NS_MEDIA is a pseudo-namespace for linking directly to a file
1761 # FIXME: Should do batch file existence checks, see comment below
1762 if( $ns == NS_MEDIA
) {
1763 wfProfileIn( __METHOD__
."-media" );
1764 # Give extensions a chance to select the file revision for us
1765 $skip = $time = false;
1766 wfRunHooks( 'BeforeParserMakeImageLinkObj', array( &$this, &$nt, &$skip, &$time ) );
1768 $link = $sk->link( $nt );
1770 $link = $sk->makeMediaLinkObj( $nt, $text, $time );
1772 # Cloak with NOPARSE to avoid replacement in replaceExternalLinks
1773 $s .= $prefix . $this->armorLinks( $link ) . $trail;
1774 $this->mOutput
->addImage( $nt->getDBkey() );
1775 wfProfileOut( __METHOD__
."-media" );
1779 wfProfileIn( __METHOD__
."-always_known" );
1780 # Some titles, such as valid special pages or files in foreign repos, should
1781 # be shown as bluelinks even though they're not included in the page table
1783 # FIXME: isAlwaysKnown() can be expensive for file links; we should really do
1784 # batch file existence checks for NS_FILE and NS_MEDIA
1785 if( $iw == '' && $nt->isAlwaysKnown() ) {
1786 $this->mOutput
->addLink( $nt );
1787 $s .= $this->makeKnownLinkHolder( $nt, $text, '', $trail, $prefix );
1789 # Links will be added to the output link list after checking
1790 $s .= $holders->makeHolder( $nt, $text, '', $trail, $prefix );
1792 wfProfileOut( __METHOD__
."-always_known" );
1794 wfProfileOut( __METHOD__
);
1799 * Make a link placeholder. The text returned can be later resolved to a real link with
1800 * replaceLinkHolders(). This is done for two reasons: firstly to avoid further
1801 * parsing of interwiki links, and secondly to allow all existence checks and
1802 * article length checks (for stub links) to be bundled into a single query.
1806 function makeLinkHolder( &$nt, $text = '', $query = '', $trail = '', $prefix = '' ) {
1807 return $this->mLinkHolders
->makeHolder( $nt, $text, $query, $trail, $prefix );
1811 * Render a forced-blue link inline; protect against double expansion of
1812 * URLs if we're in a mode that prepends full URL prefixes to internal links.
1813 * Since this little disaster has to split off the trail text to avoid
1814 * breaking URLs in the following text without breaking trails on the
1815 * wiki links, it's been made into a horrible function.
1818 * @param string $text
1819 * @param string $query
1820 * @param string $trail
1821 * @param string $prefix
1822 * @return string HTML-wikitext mix oh yuck
1824 function makeKnownLinkHolder( $nt, $text = '', $query = '', $trail = '', $prefix = '' ) {
1825 list( $inside, $trail ) = Linker
::splitTrail( $trail );
1826 $sk = $this->mOptions
->getSkin();
1827 // FIXME: use link() instead of deprecated makeKnownLinkObj()
1828 $link = $sk->makeKnownLinkObj( $nt, $text, $query, $inside, $prefix );
1829 return $this->armorLinks( $link ) . $trail;
1833 * Insert a NOPARSE hacky thing into any inline links in a chunk that's
1834 * going to go through further parsing steps before inline URL expansion.
1836 * Not needed quite as much as it used to be since free links are a bit
1837 * more sensible these days. But bracketed links are still an issue.
1839 * @param string more-or-less HTML
1840 * @return string less-or-more HTML with NOPARSE bits
1842 function armorLinks( $text ) {
1843 return preg_replace( '/\b(' . wfUrlProtocols() . ')/',
1844 "{$this->mUniqPrefix}NOPARSE$1", $text );
1848 * Return true if subpage links should be expanded on this page.
1851 function areSubpagesAllowed() {
1852 # Some namespaces don't allow subpages
1853 return MWNamespace
::hasSubpages( $this->mTitle
->getNamespace() );
1857 * Handle link to subpage if necessary
1858 * @param string $target the source of the link
1859 * @param string &$text the link text, modified as necessary
1860 * @return string the full name of the link
1863 function maybeDoSubpageLink($target, &$text) {
1864 return Linker
::normalizeSubpageLink( $this->mTitle
, $target, $text );
1868 * Used by doBlockLevels()
1871 /* private */ function closeParagraph() {
1873 if ( '' != $this->mLastSection
) {
1874 $result = '</' . $this->mLastSection
. ">\n";
1876 $this->mInPre
= false;
1877 $this->mLastSection
= '';
1880 # getCommon() returns the length of the longest common substring
1881 # of both arguments, starting at the beginning of both.
1883 /* private */ function getCommon( $st1, $st2 ) {
1884 $fl = strlen( $st1 );
1885 $shorter = strlen( $st2 );
1886 if ( $fl < $shorter ) { $shorter = $fl; }
1888 for ( $i = 0; $i < $shorter; ++
$i ) {
1889 if ( $st1{$i} != $st2{$i} ) { break; }
1893 # These next three functions open, continue, and close the list
1894 # element appropriate to the prefix character passed into them.
1896 /* private */ function openList( $char ) {
1897 $result = $this->closeParagraph();
1899 if ( '*' === $char ) { $result .= '<ul><li>'; }
1900 elseif ( '#' === $char ) { $result .= '<ol><li>'; }
1901 elseif ( ':' === $char ) { $result .= '<dl><dd>'; }
1902 elseif ( ';' === $char ) {
1903 $result .= '<dl><dt>';
1904 $this->mDTopen
= true;
1906 else { $result = '<!-- ERR 1 -->'; }
1911 /* private */ function nextItem( $char ) {
1912 if ( '*' === $char ||
'#' === $char ) { return '</li><li>'; }
1913 elseif ( ':' === $char ||
';' === $char ) {
1915 if ( $this->mDTopen
) { $close = '</dt>'; }
1916 if ( ';' === $char ) {
1917 $this->mDTopen
= true;
1918 return $close . '<dt>';
1920 $this->mDTopen
= false;
1921 return $close . '<dd>';
1924 return '<!-- ERR 2 -->';
1927 /* private */ function closeList( $char ) {
1928 if ( '*' === $char ) { $text = '</li></ul>'; }
1929 elseif ( '#' === $char ) { $text = '</li></ol>'; }
1930 elseif ( ':' === $char ) {
1931 if ( $this->mDTopen
) {
1932 $this->mDTopen
= false;
1933 $text = '</dt></dl>';
1935 $text = '</dd></dl>';
1938 else { return '<!-- ERR 3 -->'; }
1944 * Make lists from lines starting with ':', '*', '#', etc. (DBL)
1946 * @param $linestart bool whether or not this is at the start of a line.
1948 * @return string the lists rendered as HTML
1950 function doBlockLevels( $text, $linestart ) {
1951 wfProfileIn( __METHOD__
);
1953 # Parsing through the text line by line. The main thing
1954 # happening here is handling of block-level elements p, pre,
1955 # and making lists from lines starting with * # : etc.
1957 $textLines = StringUtils
::explode( "\n", $text );
1959 $lastPrefix = $output = '';
1960 $this->mDTopen
= $inBlockElem = false;
1962 $paragraphStack = false;
1964 foreach ( $textLines as $oLine ) {
1966 if ( !$linestart ) {
1976 $lastPrefixLength = strlen( $lastPrefix );
1977 $preCloseMatch = preg_match('/<\\/pre/i', $oLine );
1978 $preOpenMatch = preg_match('/<pre/i', $oLine );
1979 // If not in a <pre> element, scan for and figure out what prefixes are there.
1980 if ( !$this->mInPre
) {
1981 # Multiple prefixes may abut each other for nested lists.
1982 $prefixLength = strspn( $oLine, '*#:;' );
1983 $prefix = substr( $oLine, 0, $prefixLength );
1986 // ; and : are both from definition-lists, so they're equivalent
1987 // for the purposes of determining whether or not we need to open/close
1989 $prefix2 = str_replace( ';', ':', $prefix );
1990 $t = substr( $oLine, $prefixLength );
1991 $this->mInPre
= (bool)$preOpenMatch;
1993 # Don't interpret any other prefixes in preformatted text
1995 $prefix = $prefix2 = '';
2000 if( $prefixLength && $lastPrefix === $prefix2 ) {
2001 # Same as the last item, so no need to deal with nesting or opening stuff
2002 $output .= $this->nextItem( substr( $prefix, -1 ) );
2003 $paragraphStack = false;
2005 if ( substr( $prefix, -1 ) === ';') {
2006 # The one nasty exception: definition lists work like this:
2007 # ; title : definition text
2008 # So we check for : in the remainder text to split up the
2009 # title and definition, without b0rking links.
2011 if ($this->findColonNoLinks($t, $term, $t2) !== false) {
2013 $output .= $term . $this->nextItem( ':' );
2016 } elseif( $prefixLength ||
$lastPrefixLength ) {
2017 // We need to open or close prefixes, or both.
2019 # Either open or close a level...
2020 $commonPrefixLength = $this->getCommon( $prefix, $lastPrefix );
2021 $paragraphStack = false;
2023 // Close all the prefixes which aren't shared.
2024 while( $commonPrefixLength < $lastPrefixLength ) {
2025 $output .= $this->closeList( $lastPrefix[$lastPrefixLength-1] );
2026 --$lastPrefixLength;
2029 // Continue the current prefix if appropriate.
2030 if ( $prefixLength <= $commonPrefixLength && $commonPrefixLength > 0 ) {
2031 $output .= $this->nextItem( $prefix[$commonPrefixLength-1] );
2034 // Open prefixes where appropriate.
2035 while ( $prefixLength > $commonPrefixLength ) {
2036 $char = substr( $prefix, $commonPrefixLength, 1 );
2037 $output .= $this->openList( $char );
2039 if ( ';' === $char ) {
2040 # FIXME: This is dupe of code above
2041 if ($this->findColonNoLinks($t, $term, $t2) !== false) {
2043 $output .= $term . $this->nextItem( ':' );
2046 ++
$commonPrefixLength;
2048 $lastPrefix = $prefix2;
2051 // If we have no prefixes, go to paragraph mode.
2052 if( 0 == $prefixLength ) {
2053 wfProfileIn( __METHOD__
."-paragraph" );
2054 # No prefix (not in list)--go to paragraph mode
2055 // XXX: use a stack for nestable elements like span, table and div
2056 $openmatch = preg_match('/(?:<table|<blockquote|<h1|<h2|<h3|<h4|<h5|<h6|<pre|<tr|<p|<ul|<ol|<li|<\\/tr|<\\/td|<\\/th)/iS', $t );
2057 $closematch = preg_match(
2058 '/(?:<\\/table|<\\/blockquote|<\\/h1|<\\/h2|<\\/h3|<\\/h4|<\\/h5|<\\/h6|'.
2059 '<td|<th|<\\/?div|<hr|<\\/pre|<\\/p|'.$this->mUniqPrefix
.'-pre|<\\/li|<\\/ul|<\\/ol|<\\/?center)/iS', $t );
2060 if ( $openmatch or $closematch ) {
2061 $paragraphStack = false;
2062 #Â TODO bug 5718: paragraph closed
2063 $output .= $this->closeParagraph();
2064 if ( $preOpenMatch and !$preCloseMatch ) {
2065 $this->mInPre
= true;
2067 if ( $closematch ) {
2068 $inBlockElem = false;
2070 $inBlockElem = true;
2072 } else if ( !$inBlockElem && !$this->mInPre
) {
2073 if ( ' ' == substr( $t, 0, 1 ) and ( $this->mLastSection
=== 'pre' or trim($t) != '' ) ) {
2075 if ($this->mLastSection
!== 'pre') {
2076 $paragraphStack = false;
2077 $output .= $this->closeParagraph().'<pre>';
2078 $this->mLastSection
= 'pre';
2080 $t = substr( $t, 1 );
2083 if ( '' == trim($t) ) {
2084 if ( $paragraphStack ) {
2085 $output .= $paragraphStack.'<br />';
2086 $paragraphStack = false;
2087 $this->mLastSection
= 'p';
2089 if ($this->mLastSection
!== 'p' ) {
2090 $output .= $this->closeParagraph();
2091 $this->mLastSection
= '';
2092 $paragraphStack = '<p>';
2094 $paragraphStack = '</p><p>';
2098 if ( $paragraphStack ) {
2099 $output .= $paragraphStack;
2100 $paragraphStack = false;
2101 $this->mLastSection
= 'p';
2102 } else if ($this->mLastSection
!== 'p') {
2103 $output .= $this->closeParagraph().'<p>';
2104 $this->mLastSection
= 'p';
2109 wfProfileOut( __METHOD__
."-paragraph" );
2111 // somewhere above we forget to get out of pre block (bug 785)
2112 if($preCloseMatch && $this->mInPre
) {
2113 $this->mInPre
= false;
2115 if ($paragraphStack === false) {
2119 while ( $prefixLength ) {
2120 $output .= $this->closeList( $prefix2[$prefixLength-1] );
2123 if ( '' != $this->mLastSection
) {
2124 $output .= '</' . $this->mLastSection
. '>';
2125 $this->mLastSection
= '';
2128 wfProfileOut( __METHOD__
);
2133 * Split up a string on ':', ignoring any occurences inside tags
2134 * to prevent illegal overlapping.
2135 * @param string $str the string to split
2136 * @param string &$before set to everything before the ':'
2137 * @param string &$after set to everything after the ':'
2138 * return string the position of the ':', or false if none found
2140 function findColonNoLinks($str, &$before, &$after) {
2141 wfProfileIn( __METHOD__
);
2143 $pos = strpos( $str, ':' );
2144 if( $pos === false ) {
2146 wfProfileOut( __METHOD__
);
2150 $lt = strpos( $str, '<' );
2151 if( $lt === false ||
$lt > $pos ) {
2152 // Easy; no tag nesting to worry about
2153 $before = substr( $str, 0, $pos );
2154 $after = substr( $str, $pos+
1 );
2155 wfProfileOut( __METHOD__
);
2159 // Ugly state machine to walk through avoiding tags.
2160 $state = self
::COLON_STATE_TEXT
;
2162 $len = strlen( $str );
2163 for( $i = 0; $i < $len; $i++
) {
2167 // (Using the number is a performance hack for common cases)
2168 case 0: // self::COLON_STATE_TEXT:
2171 // Could be either a <start> tag or an </end> tag
2172 $state = self
::COLON_STATE_TAGSTART
;
2177 $before = substr( $str, 0, $i );
2178 $after = substr( $str, $i +
1 );
2179 wfProfileOut( __METHOD__
);
2182 // Embedded in a tag; don't break it.
2185 // Skip ahead looking for something interesting
2186 $colon = strpos( $str, ':', $i );
2187 if( $colon === false ) {
2188 // Nothing else interesting
2189 wfProfileOut( __METHOD__
);
2192 $lt = strpos( $str, '<', $i );
2193 if( $stack === 0 ) {
2194 if( $lt === false ||
$colon < $lt ) {
2196 $before = substr( $str, 0, $colon );
2197 $after = substr( $str, $colon +
1 );
2198 wfProfileOut( __METHOD__
);
2202 if( $lt === false ) {
2203 // Nothing else interesting to find; abort!
2204 // We're nested, but there's no close tags left. Abort!
2207 // Skip ahead to next tag start
2209 $state = self
::COLON_STATE_TAGSTART
;
2212 case 1: // self::COLON_STATE_TAG:
2217 $state = self
::COLON_STATE_TEXT
;
2220 // Slash may be followed by >?
2221 $state = self
::COLON_STATE_TAGSLASH
;
2227 case 2: // self::COLON_STATE_TAGSTART:
2230 $state = self
::COLON_STATE_CLOSETAG
;
2233 $state = self
::COLON_STATE_COMMENT
;
2236 // Illegal early close? This shouldn't happen D:
2237 $state = self
::COLON_STATE_TEXT
;
2240 $state = self
::COLON_STATE_TAG
;
2243 case 3: // self::COLON_STATE_CLOSETAG:
2248 wfDebug( __METHOD__
.": Invalid input; too many close tags\n" );
2249 wfProfileOut( __METHOD__
);
2252 $state = self
::COLON_STATE_TEXT
;
2255 case self
::COLON_STATE_TAGSLASH
:
2257 // Yes, a self-closed tag <blah/>
2258 $state = self
::COLON_STATE_TEXT
;
2260 // Probably we're jumping the gun, and this is an attribute
2261 $state = self
::COLON_STATE_TAG
;
2264 case 5: // self::COLON_STATE_COMMENT:
2266 $state = self
::COLON_STATE_COMMENTDASH
;
2269 case self
::COLON_STATE_COMMENTDASH
:
2271 $state = self
::COLON_STATE_COMMENTDASHDASH
;
2273 $state = self
::COLON_STATE_COMMENT
;
2276 case self
::COLON_STATE_COMMENTDASHDASH
:
2278 $state = self
::COLON_STATE_TEXT
;
2280 $state = self
::COLON_STATE_COMMENT
;
2284 throw new MWException( "State machine error in " . __METHOD__
);
2288 wfDebug( __METHOD__
.": Invalid input; not enough close tags (stack $stack, state $state)\n" );
2291 wfProfileOut( __METHOD__
);
2296 * Return value of a magic variable (like PAGENAME)
2300 function getVariableValue( $index, $frame=false ) {
2301 global $wgContLang, $wgSitename, $wgServer, $wgServerName;
2302 global $wgScriptPath, $wgStylePath;
2305 * Some of these require message or data lookups and can be
2306 * expensive to check many times.
2308 if ( wfRunHooks( 'ParserGetVariableValueVarCache', array( &$this, &$this->mVarCache
) ) ) {
2309 if ( isset( $this->mVarCache
[$index] ) ) {
2310 return $this->mVarCache
[$index];
2314 $ts = wfTimestamp( TS_UNIX
, $this->mOptions
->getTimestamp() );
2315 wfRunHooks( 'ParserGetVariableValueTs', array( &$this, &$ts ) );
2318 global $wgLocaltimezone;
2319 if ( isset( $wgLocaltimezone ) ) {
2320 $oldtz = getenv( 'TZ' );
2321 putenv( 'TZ='.$wgLocaltimezone );
2324 wfSuppressWarnings(); // E_STRICT system time bitching
2325 $localTimestamp = date( 'YmdHis', $ts );
2326 $localMonth = date( 'm', $ts );
2327 $localMonth1 = date( 'n', $ts );
2328 $localMonthName = date( 'n', $ts );
2329 $localDay = date( 'j', $ts );
2330 $localDay2 = date( 'd', $ts );
2331 $localDayOfWeek = date( 'w', $ts );
2332 $localWeek = date( 'W', $ts );
2333 $localYear = date( 'Y', $ts );
2334 $localHour = date( 'H', $ts );
2335 if ( isset( $wgLocaltimezone ) ) {
2336 putenv( 'TZ='.$oldtz );
2338 wfRestoreWarnings();
2341 case 'currentmonth':
2342 return $this->mVarCache
[$index] = $wgContLang->formatNum( gmdate( 'm', $ts ) );
2343 case 'currentmonth1':
2344 return $this->mVarCache
[$index] = $wgContLang->formatNum( gmdate( 'n', $ts ) );
2345 case 'currentmonthname':
2346 return $this->mVarCache
[$index] = $wgContLang->getMonthName( gmdate( 'n', $ts ) );
2347 case 'currentmonthnamegen':
2348 return $this->mVarCache
[$index] = $wgContLang->getMonthNameGen( gmdate( 'n', $ts ) );
2349 case 'currentmonthabbrev':
2350 return $this->mVarCache
[$index] = $wgContLang->getMonthAbbreviation( gmdate( 'n', $ts ) );
2352 return $this->mVarCache
[$index] = $wgContLang->formatNum( gmdate( 'j', $ts ) );
2354 return $this->mVarCache
[$index] = $wgContLang->formatNum( gmdate( 'd', $ts ) );
2356 return $this->mVarCache
[$index] = $wgContLang->formatNum( $localMonth );
2358 return $this->mVarCache
[$index] = $wgContLang->formatNum( $localMonth1 );
2359 case 'localmonthname':
2360 return $this->mVarCache
[$index] = $wgContLang->getMonthName( $localMonthName );
2361 case 'localmonthnamegen':
2362 return $this->mVarCache
[$index] = $wgContLang->getMonthNameGen( $localMonthName );
2363 case 'localmonthabbrev':
2364 return $this->mVarCache
[$index] = $wgContLang->getMonthAbbreviation( $localMonthName );
2366 return $this->mVarCache
[$index] = $wgContLang->formatNum( $localDay );
2368 return $this->mVarCache
[$index] = $wgContLang->formatNum( $localDay2 );
2370 return wfEscapeWikiText( $this->mTitle
->getText() );
2372 return $this->mTitle
->getPartialURL();
2373 case 'fullpagename':
2374 return wfEscapeWikiText( $this->mTitle
->getPrefixedText() );
2375 case 'fullpagenamee':
2376 return $this->mTitle
->getPrefixedURL();
2378 return wfEscapeWikiText( $this->mTitle
->getSubpageText() );
2379 case 'subpagenamee':
2380 return $this->mTitle
->getSubpageUrlForm();
2381 case 'basepagename':
2382 return wfEscapeWikiText( $this->mTitle
->getBaseText() );
2383 case 'basepagenamee':
2384 return wfUrlEncode( str_replace( ' ', '_', $this->mTitle
->getBaseText() ) );
2385 case 'talkpagename':
2386 if( $this->mTitle
->canTalk() ) {
2387 $talkPage = $this->mTitle
->getTalkPage();
2388 return wfEscapeWikiText( $talkPage->getPrefixedText() );
2392 case 'talkpagenamee':
2393 if( $this->mTitle
->canTalk() ) {
2394 $talkPage = $this->mTitle
->getTalkPage();
2395 return $talkPage->getPrefixedUrl();
2399 case 'subjectpagename':
2400 $subjPage = $this->mTitle
->getSubjectPage();
2401 return wfEscapeWikiText( $subjPage->getPrefixedText() );
2402 case 'subjectpagenamee':
2403 $subjPage = $this->mTitle
->getSubjectPage();
2404 return $subjPage->getPrefixedUrl();
2406 // Let the edit saving system know we should parse the page
2407 // *after* a revision ID has been assigned.
2408 $this->mOutput
->setFlag( 'vary-revision' );
2409 wfDebug( __METHOD__
. ": {{REVISIONID}} used, setting vary-revision...\n" );
2410 return $this->mRevisionId
;
2412 // Let the edit saving system know we should parse the page
2413 // *after* a revision ID has been assigned. This is for null edits.
2414 $this->mOutput
->setFlag( 'vary-revision' );
2415 wfDebug( __METHOD__
. ": {{REVISIONDAY}} used, setting vary-revision...\n" );
2416 return intval( substr( $this->getRevisionTimestamp(), 6, 2 ) );
2417 case 'revisionday2':
2418 // Let the edit saving system know we should parse the page
2419 // *after* a revision ID has been assigned. This is for null edits.
2420 $this->mOutput
->setFlag( 'vary-revision' );
2421 wfDebug( __METHOD__
. ": {{REVISIONDAY2}} used, setting vary-revision...\n" );
2422 return substr( $this->getRevisionTimestamp(), 6, 2 );
2423 case 'revisionmonth':
2424 // Let the edit saving system know we should parse the page
2425 // *after* a revision ID has been assigned. This is for null edits.
2426 $this->mOutput
->setFlag( 'vary-revision' );
2427 wfDebug( __METHOD__
. ": {{REVISIONMONTH}} used, setting vary-revision...\n" );
2428 return intval( substr( $this->getRevisionTimestamp(), 4, 2 ) );
2429 case 'revisionyear':
2430 // Let the edit saving system know we should parse the page
2431 // *after* a revision ID has been assigned. This is for null edits.
2432 $this->mOutput
->setFlag( 'vary-revision' );
2433 wfDebug( __METHOD__
. ": {{REVISIONYEAR}} used, setting vary-revision...\n" );
2434 return substr( $this->getRevisionTimestamp(), 0, 4 );
2435 case 'revisiontimestamp':
2436 // Let the edit saving system know we should parse the page
2437 // *after* a revision ID has been assigned. This is for null edits.
2438 $this->mOutput
->setFlag( 'vary-revision' );
2439 wfDebug( __METHOD__
. ": {{REVISIONTIMESTAMP}} used, setting vary-revision...\n" );
2440 return $this->getRevisionTimestamp();
2441 case 'revisionuser':
2442 // Let the edit saving system know we should parse the page
2443 // *after* a revision ID has been assigned. This is for null edits.
2444 $this->mOutput
->setFlag( 'vary-revision' );
2445 wfDebug( __METHOD__
. ": {{REVISIONUSER}} used, setting vary-revision...\n" );
2446 return $this->getRevisionUser();
2448 return str_replace('_',' ',$wgContLang->getNsText( $this->mTitle
->getNamespace() ) );
2450 return wfUrlencode( $wgContLang->getNsText( $this->mTitle
->getNamespace() ) );
2452 return $this->mTitle
->canTalk() ?
str_replace('_',' ',$this->mTitle
->getTalkNsText()) : '';
2454 return $this->mTitle
->canTalk() ?
wfUrlencode( $this->mTitle
->getTalkNsText() ) : '';
2455 case 'subjectspace':
2456 return $this->mTitle
->getSubjectNsText();
2457 case 'subjectspacee':
2458 return( wfUrlencode( $this->mTitle
->getSubjectNsText() ) );
2459 case 'currentdayname':
2460 return $this->mVarCache
[$index] = $wgContLang->getWeekdayName( gmdate( 'w', $ts ) +
1 );
2462 return $this->mVarCache
[$index] = $wgContLang->formatNum( gmdate( 'Y', $ts ), true );
2464 return $this->mVarCache
[$index] = $wgContLang->time( wfTimestamp( TS_MW
, $ts ), false, false );
2466 return $this->mVarCache
[$index] = $wgContLang->formatNum( gmdate( 'H', $ts ), true );
2468 // @bug 4594 PHP5 has it zero padded, PHP4 does not, cast to
2469 // int to remove the padding
2470 return $this->mVarCache
[$index] = $wgContLang->formatNum( (int)gmdate( 'W', $ts ) );
2472 return $this->mVarCache
[$index] = $wgContLang->formatNum( gmdate( 'w', $ts ) );
2473 case 'localdayname':
2474 return $this->mVarCache
[$index] = $wgContLang->getWeekdayName( $localDayOfWeek +
1 );
2476 return $this->mVarCache
[$index] = $wgContLang->formatNum( $localYear, true );
2478 return $this->mVarCache
[$index] = $wgContLang->time( $localTimestamp, false, false );
2480 return $this->mVarCache
[$index] = $wgContLang->formatNum( $localHour, true );
2482 // @bug 4594 PHP5 has it zero padded, PHP4 does not, cast to
2483 // int to remove the padding
2484 return $this->mVarCache
[$index] = $wgContLang->formatNum( (int)$localWeek );
2486 return $this->mVarCache
[$index] = $wgContLang->formatNum( $localDayOfWeek );
2487 case 'numberofarticles':
2488 return $this->mVarCache
[$index] = $wgContLang->formatNum( SiteStats
::articles() );
2489 case 'numberoffiles':
2490 return $this->mVarCache
[$index] = $wgContLang->formatNum( SiteStats
::images() );
2491 case 'numberofusers':
2492 return $this->mVarCache
[$index] = $wgContLang->formatNum( SiteStats
::users() );
2493 case 'numberofactiveusers':
2494 return $this->mVarCache
[$index] = $wgContLang->formatNum( SiteStats
::activeUsers() );
2495 case 'numberofpages':
2496 return $this->mVarCache
[$index] = $wgContLang->formatNum( SiteStats
::pages() );
2497 case 'numberofadmins':
2498 return $this->mVarCache
[$index] = $wgContLang->formatNum( SiteStats
::numberingroup('sysop') );
2499 case 'numberofedits':
2500 return $this->mVarCache
[$index] = $wgContLang->formatNum( SiteStats
::edits() );
2501 case 'numberofviews':
2502 return $this->mVarCache
[$index] = $wgContLang->formatNum( SiteStats
::views() );
2503 case 'currenttimestamp':
2504 return $this->mVarCache
[$index] = wfTimestamp( TS_MW
, $ts );
2505 case 'localtimestamp':
2506 return $this->mVarCache
[$index] = $localTimestamp;
2507 case 'currentversion':
2508 return $this->mVarCache
[$index] = SpecialVersion
::getVersion();
2514 return $wgServerName;
2516 return $wgScriptPath;
2518 return $wgStylePath;
2519 case 'directionmark':
2520 return $wgContLang->getDirMark();
2521 case 'contentlanguage':
2522 global $wgContLanguageCode;
2523 return $wgContLanguageCode;
2526 if ( wfRunHooks( 'ParserGetVariableValueSwitch', array( &$this, &$this->mVarCache
, &$index, &$ret, &$frame ) ) )
2534 * initialise the magic variables (like CURRENTMONTHNAME)
2538 function initialiseVariables() {
2539 wfProfileIn( __METHOD__
);
2540 $variableIDs = MagicWord
::getVariableIDs();
2542 $this->mVariables
= new MagicWordArray( $variableIDs );
2543 wfProfileOut( __METHOD__
);
2547 * Preprocess some wikitext and return the document tree.
2548 * This is the ghost of replace_variables().
2550 * @param string $text The text to parse
2551 * @param integer flags Bitwise combination of:
2552 * self::PTD_FOR_INCLUSION Handle <noinclude>/<includeonly> as if the text is being
2553 * included. Default is to assume a direct page view.
2555 * The generated DOM tree must depend only on the input text and the flags.
2556 * The DOM tree must be the same in OT_HTML and OT_WIKI mode, to avoid a regression of bug 4899.
2558 * Any flag added to the $flags parameter here, or any other parameter liable to cause a
2559 * change in the DOM tree for a given text, must be passed through the section identifier
2560 * in the section edit link and thus back to extractSections().
2562 * The output of this function is currently only cached in process memory, but a persistent
2563 * cache may be implemented at a later date which takes further advantage of these strict
2564 * dependency requirements.
2568 function preprocessToDom ( $text, $flags = 0 ) {
2569 $dom = $this->getPreprocessor()->preprocessToObj( $text, $flags );
2574 * Return a three-element array: leading whitespace, string contents, trailing whitespace
2576 public static function splitWhitespace( $s ) {
2577 $ltrimmed = ltrim( $s );
2578 $w1 = substr( $s, 0, strlen( $s ) - strlen( $ltrimmed ) );
2579 $trimmed = rtrim( $ltrimmed );
2580 $diff = strlen( $ltrimmed ) - strlen( $trimmed );
2582 $w2 = substr( $ltrimmed, -$diff );
2586 return array( $w1, $trimmed, $w2 );
2590 * Replace magic variables, templates, and template arguments
2591 * with the appropriate text. Templates are substituted recursively,
2592 * taking care to avoid infinite loops.
2594 * Note that the substitution depends on value of $mOutputType:
2595 * self::OT_WIKI: only {{subst:}} templates
2596 * self::OT_PREPROCESS: templates but not extension tags
2597 * self::OT_HTML: all templates and extension tags
2599 * @param string $tex The text to transform
2600 * @param PPFrame $frame Object describing the arguments passed to the template.
2601 * Arguments may also be provided as an associative array, as was the usual case before MW1.12.
2602 * Providing arguments this way may be useful for extensions wishing to perform variable replacement explicitly.
2603 * @param bool $argsOnly Only do argument (triple-brace) expansion, not double-brace expansion
2606 function replaceVariables( $text, $frame = false, $argsOnly = false ) {
2607 # Is there any text? Also, Prevent too big inclusions!
2608 if ( strlen( $text ) < 1 ||
strlen( $text ) > $this->mOptions
->getMaxIncludeSize() ) {
2611 wfProfileIn( __METHOD__
);
2613 if ( $frame === false ) {
2614 $frame = $this->getPreprocessor()->newFrame();
2615 } elseif ( !( $frame instanceof PPFrame
) ) {
2616 wfDebug( __METHOD__
." called using plain parameters instead of a PPFrame instance. Creating custom frame.\n" );
2617 $frame = $this->getPreprocessor()->newCustomFrame($frame);
2620 $dom = $this->preprocessToDom( $text );
2621 $flags = $argsOnly ? PPFrame
::NO_TEMPLATES
: 0;
2622 $text = $frame->expand( $dom, $flags );
2624 wfProfileOut( __METHOD__
);
2628 /// Clean up argument array - refactored in 1.9 so parserfunctions can use it, too.
2629 static function createAssocArgs( $args ) {
2630 $assocArgs = array();
2632 foreach( $args as $arg ) {
2633 $eqpos = strpos( $arg, '=' );
2634 if ( $eqpos === false ) {
2635 $assocArgs[$index++
] = $arg;
2637 $name = trim( substr( $arg, 0, $eqpos ) );
2638 $value = trim( substr( $arg, $eqpos+
1 ) );
2639 if ( $value === false ) {
2642 if ( $name !== false ) {
2643 $assocArgs[$name] = $value;
2652 * Warn the user when a parser limitation is reached
2653 * Will warn at most once the user per limitation type
2655 * @param string $limitationType, should be one of:
2656 * 'expensive-parserfunction' (corresponding messages: 'expensive-parserfunction-warning', 'expensive-parserfunction-category')
2657 * 'post-expand-template-argument' (corresponding messages: 'post-expand-template-argument-warning', 'post-expand-template-argument-category')
2658 * 'post-expand-template-inclusion' (corresponding messages: 'post-expand-template-inclusion-warning', 'post-expand-template-inclusion-category')
2659 * @params int $current, $max When an explicit limit has been
2660 * exceeded, provide the values (optional)
2662 function limitationWarn( $limitationType, $current=null, $max=null) {
2663 $msgName = $limitationType . '-warning';
2664 //does no harm if $current and $max are present but are unnecessary for the message
2665 $warning = wfMsgExt( $msgName, array( 'parsemag', 'escape' ), $current, $max );
2666 $this->mOutput
->addWarning( $warning );
2667 $cat = Title
::makeTitleSafe( NS_CATEGORY
, wfMsgForContent( $limitationType . '-category' ) );
2669 $this->mOutput
->addCategory( $cat->getDBkey(), $this->getDefaultSort() );
2674 * Return the text of a template, after recursively
2675 * replacing any variables or templates within the template.
2677 * @param array $piece The parts of the template
2678 * $piece['title']: the title, i.e. the part before the |
2679 * $piece['parts']: the parameter array
2680 * $piece['lineStart']: whether the brace was at the start of a line
2681 * @param PPFrame The current frame, contains template arguments
2682 * @return string the text of the template
2685 function braceSubstitution( $piece, $frame ) {
2686 global $wgContLang, $wgNonincludableNamespaces;
2687 wfProfileIn( __METHOD__
);
2688 wfProfileIn( __METHOD__
.'-setup' );
2691 $found = false; # $text has been filled
2692 $nowiki = false; # wiki markup in $text should be escaped
2693 $isHTML = false; # $text is HTML, armour it against wikitext transformation
2694 $forceRawInterwiki = false; # Force interwiki transclusion to be done in raw mode not rendered
2695 $isChildObj = false; # $text is a DOM node needing expansion in a child frame
2696 $isLocalObj = false; # $text is a DOM node needing expansion in the current frame
2698 # Title object, where $text came from
2701 # $part1 is the bit before the first |, and must contain only title characters.
2702 # Various prefixes will be stripped from it later.
2703 $titleWithSpaces = $frame->expand( $piece['title'] );
2704 $part1 = trim( $titleWithSpaces );
2707 # Original title text preserved for various purposes
2708 $originalTitle = $part1;
2710 # $args is a list of argument nodes, starting from index 0, not including $part1
2711 $args = (null == $piece['parts']) ?
array() : $piece['parts'];
2712 wfProfileOut( __METHOD__
.'-setup' );
2715 wfProfileIn( __METHOD__
.'-modifiers' );
2717 $mwSubst = MagicWord
::get( 'subst' );
2718 if ( $mwSubst->matchStartAndRemove( $part1 ) xor $this->ot
['wiki'] ) {
2719 # One of two possibilities is true:
2720 # 1) Found SUBST but not in the PST phase
2721 # 2) Didn't find SUBST and in the PST phase
2722 # In either case, return without further processing
2723 $text = $frame->virtualBracketedImplode( '{{', '|', '}}', $titleWithSpaces, $args );
2730 if ( !$found && $args->getLength() == 0 ) {
2731 $id = $this->mVariables
->matchStartToEnd( $part1 );
2732 if ( $id !== false ) {
2733 $text = $this->getVariableValue( $id, $frame );
2734 if (MagicWord
::getCacheTTL($id)>-1)
2735 $this->mOutput
->mContainsOldMagic
= true;
2740 # MSG, MSGNW and RAW
2743 $mwMsgnw = MagicWord
::get( 'msgnw' );
2744 if ( $mwMsgnw->matchStartAndRemove( $part1 ) ) {
2747 # Remove obsolete MSG:
2748 $mwMsg = MagicWord
::get( 'msg' );
2749 $mwMsg->matchStartAndRemove( $part1 );
2753 $mwRaw = MagicWord
::get( 'raw' );
2754 if ( $mwRaw->matchStartAndRemove( $part1 ) ) {
2755 $forceRawInterwiki = true;
2758 wfProfileOut( __METHOD__
.'-modifiers' );
2762 wfProfileIn( __METHOD__
. '-pfunc' );
2764 $colonPos = strpos( $part1, ':' );
2765 if ( $colonPos !== false ) {
2766 # Case sensitive functions
2767 $function = substr( $part1, 0, $colonPos );
2768 if ( isset( $this->mFunctionSynonyms
[1][$function] ) ) {
2769 $function = $this->mFunctionSynonyms
[1][$function];
2771 # Case insensitive functions
2772 $function = $wgContLang->lc( $function );
2773 if ( isset( $this->mFunctionSynonyms
[0][$function] ) ) {
2774 $function = $this->mFunctionSynonyms
[0][$function];
2780 list( $callback, $flags ) = $this->mFunctionHooks
[$function];
2781 $initialArgs = array( &$this );
2782 $funcArgs = array( trim( substr( $part1, $colonPos +
1 ) ) );
2783 if ( $flags & SFH_OBJECT_ARGS
) {
2784 # Add a frame parameter, and pass the arguments as an array
2785 $allArgs = $initialArgs;
2786 $allArgs[] = $frame;
2787 for ( $i = 0; $i < $args->getLength(); $i++
) {
2788 $funcArgs[] = $args->item( $i );
2790 $allArgs[] = $funcArgs;
2792 # Convert arguments to plain text
2793 for ( $i = 0; $i < $args->getLength(); $i++
) {
2794 $funcArgs[] = trim( $frame->expand( $args->item( $i ) ) );
2796 $allArgs = array_merge( $initialArgs, $funcArgs );
2799 # Workaround for PHP bug 35229 and similar
2800 if ( !is_callable( $callback ) ) {
2801 wfProfileOut( __METHOD__
. '-pfunc' );
2802 wfProfileOut( __METHOD__
);
2803 throw new MWException( "Tag hook for $function is not callable\n" );
2805 $result = call_user_func_array( $callback, $allArgs );
2808 $preprocessFlags = 0;
2810 if ( is_array( $result ) ) {
2811 if ( isset( $result[0] ) ) {
2813 unset( $result[0] );
2816 // Extract flags into the local scope
2817 // This allows callers to set flags such as nowiki, found, etc.
2823 $text = $this->preprocessToDom( $text, $preprocessFlags );
2828 wfProfileOut( __METHOD__
. '-pfunc' );
2831 # Finish mangling title and then check for loops.
2832 # Set $title to a Title object and $titleText to the PDBK
2835 # Split the title into page and subpage
2837 $part1 = $this->maybeDoSubpageLink( $part1, $subpage );
2838 if ($subpage !== '') {
2839 $ns = $this->mTitle
->getNamespace();
2841 $title = Title
::newFromText( $part1, $ns );
2843 $titleText = $title->getPrefixedText();
2844 # Check for language variants if the template is not found
2845 if($wgContLang->hasVariants() && $title->getArticleID() == 0){
2846 $wgContLang->findVariantLink( $part1, $title, true );
2848 # Do recursion depth check
2849 $limit = $this->mOptions
->getMaxTemplateDepth();
2850 if ( $frame->depth
>= $limit ) {
2852 $text = '<span class="error">' . wfMsgForContent( 'parser-template-recursion-depth-warning', $limit ) . '</span>';
2857 # Load from database
2858 if ( !$found && $title ) {
2859 wfProfileIn( __METHOD__
. '-loadtpl' );
2860 if ( !$title->isExternal() ) {
2861 if ( $title->getNamespace() == NS_SPECIAL
&& $this->mOptions
->getAllowSpecialInclusion() && $this->ot
['html'] ) {
2862 $text = SpecialPage
::capturePath( $title );
2863 if ( is_string( $text ) ) {
2866 $this->disableCache();
2868 } else if ( $wgNonincludableNamespaces && in_array( $title->getNamespace(), $wgNonincludableNamespaces ) ) {
2869 $found = false; //access denied
2870 wfDebug( __METHOD__
.": template inclusion denied for " . $title->getPrefixedDBkey() );
2872 list( $text, $title ) = $this->getTemplateDom( $title );
2873 if ( $text !== false ) {
2879 # If the title is valid but undisplayable, make a link to it
2880 if ( !$found && ( $this->ot
['html'] ||
$this->ot
['pre'] ) ) {
2881 $text = "[[:$titleText]]";
2884 } elseif ( $title->isTrans() ) {
2885 // Interwiki transclusion
2886 if ( $this->ot
['html'] && !$forceRawInterwiki ) {
2887 $text = $this->interwikiTransclude( $title, 'render' );
2890 $text = $this->interwikiTransclude( $title, 'raw' );
2891 // Preprocess it like a template
2892 $text = $this->preprocessToDom( $text, self
::PTD_FOR_INCLUSION
);
2898 # Do infinite loop check
2899 # This has to be done after redirect resolution to avoid infinite loops via redirects
2900 if ( !$frame->loopCheck( $title ) ) {
2902 $text = '<span class="error">' . wfMsgForContent( 'parser-template-loop-warning', $titleText ) . '</span>';
2903 wfDebug( __METHOD__
.": template loop broken at '$titleText'\n" );
2905 wfProfileOut( __METHOD__
. '-loadtpl' );
2908 # If we haven't found text to substitute by now, we're done
2909 # Recover the source wikitext and return it
2911 $text = $frame->virtualBracketedImplode( '{{', '|', '}}', $titleWithSpaces, $args );
2912 wfProfileOut( __METHOD__
);
2913 return array( 'object' => $text );
2916 # Expand DOM-style return values in a child frame
2917 if ( $isChildObj ) {
2918 # Clean up argument array
2919 $newFrame = $frame->newChild( $args, $title );
2922 $text = $newFrame->expand( $text, PPFrame
::RECOVER_ORIG
);
2923 } elseif ( $titleText !== false && $newFrame->isEmpty() ) {
2924 # Expansion is eligible for the empty-frame cache
2925 if ( isset( $this->mTplExpandCache
[$titleText] ) ) {
2926 $text = $this->mTplExpandCache
[$titleText];
2928 $text = $newFrame->expand( $text );
2929 $this->mTplExpandCache
[$titleText] = $text;
2932 # Uncached expansion
2933 $text = $newFrame->expand( $text );
2936 if ( $isLocalObj && $nowiki ) {
2937 $text = $frame->expand( $text, PPFrame
::RECOVER_ORIG
);
2938 $isLocalObj = false;
2941 # Replace raw HTML by a placeholder
2942 # Add a blank line preceding, to prevent it from mucking up
2943 # immediately preceding headings
2945 $text = "\n\n" . $this->insertStripItem( $text );
2947 # Escape nowiki-style return values
2948 elseif ( $nowiki && ( $this->ot
['html'] ||
$this->ot
['pre'] ) ) {
2949 $text = wfEscapeWikiText( $text );
2951 # Bug 529: if the template begins with a table or block-level
2952 # element, it should be treated as beginning a new line.
2953 # This behaviour is somewhat controversial.
2954 elseif ( is_string( $text ) && !$piece['lineStart'] && preg_match('/^(?:{\\||:|;|#|\*)/', $text)) /*}*/{
2955 $text = "\n" . $text;
2958 if ( is_string( $text ) && !$this->incrementIncludeSize( 'post-expand', strlen( $text ) ) ) {
2959 # Error, oversize inclusion
2960 $text = "[[$originalTitle]]" .
2961 $this->insertStripItem( '<!-- WARNING: template omitted, post-expand include size too large -->' );
2962 $this->limitationWarn( 'post-expand-template-inclusion' );
2965 if ( $isLocalObj ) {
2966 $ret = array( 'object' => $text );
2968 $ret = array( 'text' => $text );
2971 wfProfileOut( __METHOD__
);
2976 * Get the semi-parsed DOM representation of a template with a given title,
2977 * and its redirect destination title. Cached.
2979 function getTemplateDom( $title ) {
2980 $cacheTitle = $title;
2981 $titleText = $title->getPrefixedDBkey();
2983 if ( isset( $this->mTplRedirCache
[$titleText] ) ) {
2984 list( $ns, $dbk ) = $this->mTplRedirCache
[$titleText];
2985 $title = Title
::makeTitle( $ns, $dbk );
2986 $titleText = $title->getPrefixedDBkey();
2988 if ( isset( $this->mTplDomCache
[$titleText] ) ) {
2989 return array( $this->mTplDomCache
[$titleText], $title );
2992 // Cache miss, go to the database
2993 list( $text, $title ) = $this->fetchTemplateAndTitle( $title );
2995 if ( $text === false ) {
2996 $this->mTplDomCache
[$titleText] = false;
2997 return array( false, $title );
3000 $dom = $this->preprocessToDom( $text, self
::PTD_FOR_INCLUSION
);
3001 $this->mTplDomCache
[ $titleText ] = $dom;
3003 if (! $title->equals($cacheTitle)) {
3004 $this->mTplRedirCache
[$cacheTitle->getPrefixedDBkey()] =
3005 array( $title->getNamespace(),$cdb = $title->getDBkey() );
3008 return array( $dom, $title );
3012 * Fetch the unparsed text of a template and register a reference to it.
3014 function fetchTemplateAndTitle( $title ) {
3015 $templateCb = $this->mOptions
->getTemplateCallback();
3016 $stuff = call_user_func( $templateCb, $title, $this );
3017 $text = $stuff['text'];
3018 $finalTitle = isset( $stuff['finalTitle'] ) ?
$stuff['finalTitle'] : $title;
3019 if ( isset( $stuff['deps'] ) ) {
3020 foreach ( $stuff['deps'] as $dep ) {
3021 $this->mOutput
->addTemplate( $dep['title'], $dep['page_id'], $dep['rev_id'] );
3024 return array($text,$finalTitle);
3027 function fetchTemplate( $title ) {
3028 $rv = $this->fetchTemplateAndTitle($title);
3033 * Static function to get a template
3034 * Can be overridden via ParserOptions::setTemplateCallback().
3036 static function statelessFetchTemplate( $title, $parser=false ) {
3037 $text = $skip = false;
3038 $finalTitle = $title;
3041 // Loop to fetch the article, with up to 1 redirect
3042 for ( $i = 0; $i < 2 && is_object( $title ); $i++
) {
3043 # Give extensions a chance to select the revision instead
3044 $id = false; // Assume current
3045 wfRunHooks( 'BeforeParserFetchTemplateAndtitle', array( $parser, &$title, &$skip, &$id ) );
3051 'page_id' => $title->getArticleID(),
3055 $rev = $id ? Revision
::newFromId( $id ) : Revision
::newFromTitle( $title );
3056 $rev_id = $rev ?
$rev->getId() : 0;
3057 // If there is no current revision, there is no page
3058 if( $id === false && !$rev ) {
3059 $linkCache = LinkCache
::singleton();
3060 $linkCache->addBadLinkObj( $title );
3065 'page_id' => $title->getArticleID(),
3066 'rev_id' => $rev_id );
3069 $text = $rev->getText();
3070 } elseif( $title->getNamespace() == NS_MEDIAWIKI
) {
3072 $message = $wgContLang->lcfirst( $title->getText() );
3073 $text = wfMsgForContentNoTrans( $message );
3074 if( wfEmptyMsg( $message, $text ) ) {
3081 if ( $text === false ) {
3085 $finalTitle = $title;
3086 $title = Title
::newFromRedirect( $text );
3090 'finalTitle' => $finalTitle,
3095 * Transclude an interwiki link.
3097 function interwikiTransclude( $title, $action ) {
3098 global $wgEnableScaryTranscluding;
3100 if (!$wgEnableScaryTranscluding)
3101 return wfMsg('scarytranscludedisabled');
3103 $url = $title->getFullUrl( "action=$action" );
3105 if (strlen($url) > 255)
3106 return wfMsg('scarytranscludetoolong');
3107 return $this->fetchScaryTemplateMaybeFromCache($url);
3110 function fetchScaryTemplateMaybeFromCache($url) {
3111 global $wgTranscludeCacheExpiry;
3112 $dbr = wfGetDB(DB_SLAVE
);
3113 $obj = $dbr->selectRow('transcache', array('tc_time', 'tc_contents'),
3114 array('tc_url' => $url));
3116 $time = $obj->tc_time
;
3117 $text = $obj->tc_contents
;
3118 if ($time && time() < $time +
$wgTranscludeCacheExpiry ) {
3123 $text = Http
::get($url);
3125 return wfMsg('scarytranscludefailed', $url);
3127 $dbw = wfGetDB(DB_MASTER
);
3128 $dbw->replace('transcache', array('tc_url'), array(
3130 'tc_time' => time(),
3131 'tc_contents' => $text));
3137 * Triple brace replacement -- used for template arguments
3140 function argSubstitution( $piece, $frame ) {
3141 wfProfileIn( __METHOD__
);
3144 $parts = $piece['parts'];
3145 $nameWithSpaces = $frame->expand( $piece['title'] );
3146 $argName = trim( $nameWithSpaces );
3148 $text = $frame->getArgument( $argName );
3149 if ( $text === false && $parts->getLength() > 0
3153 ||
( $this->ot
['wiki'] && $frame->isTemplate() )
3156 # No match in frame, use the supplied default
3157 $object = $parts->item( 0 )->getChildren();
3159 if ( !$this->incrementIncludeSize( 'arg', strlen( $text ) ) ) {
3160 $error = '<!-- WARNING: argument omitted, expansion size too large -->';
3161 $this->limitationWarn( 'post-expand-template-argument' );
3164 if ( $text === false && $object === false ) {
3166 $object = $frame->virtualBracketedImplode( '{{{', '|', '}}}', $nameWithSpaces, $parts );
3168 if ( $error !== false ) {
3171 if ( $object !== false ) {
3172 $ret = array( 'object' => $object );
3174 $ret = array( 'text' => $text );
3177 wfProfileOut( __METHOD__
);
3182 * Return the text to be used for a given extension tag.
3183 * This is the ghost of strip().
3185 * @param array $params Associative array of parameters:
3186 * name PPNode for the tag name
3187 * attr PPNode for unparsed text where tag attributes are thought to be
3188 * attributes Optional associative array of parsed attributes
3189 * inner Contents of extension element
3190 * noClose Original text did not have a close tag
3191 * @param PPFrame $frame
3193 function extensionSubstitution( $params, $frame ) {
3194 global $wgRawHtml, $wgContLang;
3196 $name = $frame->expand( $params['name'] );
3197 $attrText = !isset( $params['attr'] ) ?
null : $frame->expand( $params['attr'] );
3198 $content = !isset( $params['inner'] ) ?
null : $frame->expand( $params['inner'] );
3200 $marker = "{$this->mUniqPrefix}-$name-" . sprintf('%08X', $this->mMarkerIndex++
) . self
::MARKER_SUFFIX
;
3202 $isFunctionTag = isset( $this->mFunctionTagHooks
[strtolower($name)] ) &&
3203 ( $this->ot
['html'] ||
$this->ot
['pre'] );
3204 if ( $this->ot
['html'] ||
$isFunctionTag ) {
3205 $name = strtolower( $name );
3206 $attributes = Sanitizer
::decodeTagAttributes( $attrText );
3207 if ( isset( $params['attributes'] ) ) {
3208 $attributes = $attributes +
$params['attributes'];
3216 throw new MWException( '<html> extension tag encountered unexpectedly' );
3219 $content = strtr($content, array('-{' => '-{', '}-' => '}-'));
3220 $output = Xml
::escapeTagsOnly( $content );
3223 $output = $wgContLang->armourMath(
3224 MathRenderer
::renderMath( $content, $attributes ) );
3227 $output = $this->renderImageGallery( $content, $attributes );
3230 if( isset( $this->mTagHooks
[$name] ) ) {
3231 # Workaround for PHP bug 35229 and similar
3232 if ( !is_callable( $this->mTagHooks
[$name] ) ) {
3233 throw new MWException( "Tag hook for $name is not callable\n" );
3235 $output = call_user_func_array( $this->mTagHooks
[$name],
3236 array( $content, $attributes, $this, $frame ) );
3237 } elseif( isset( $this->mFunctionTagHooks
[$name] ) ) {
3238 list( $callback, $flags ) = $this->mFunctionTagHooks
[$name];
3239 if( !is_callable( $callback ) )
3240 throw new MWException( "Tag hook for $name is not callable\n" );
3242 $output = call_user_func_array( $callback,
3243 array( &$this, $frame, $content, $attributes ) );
3245 $output = '<span class="error">Invalid tag extension name: ' .
3246 htmlspecialchars( $name ) . '</span>';
3250 if ( is_null( $attrText ) ) {
3253 if ( isset( $params['attributes'] ) ) {
3254 foreach ( $params['attributes'] as $attrName => $attrValue ) {
3255 $attrText .= ' ' . htmlspecialchars( $attrName ) . '="' .
3256 htmlspecialchars( $attrValue ) . '"';
3259 if ( $content === null ) {
3260 $output = "<$name$attrText/>";
3262 $close = is_null( $params['close'] ) ?
'' : $frame->expand( $params['close'] );
3263 $output = "<$name$attrText>$content$close";
3267 if( $isFunctionTag ) {
3269 } elseif ( $name === 'html' ||
$name === 'nowiki' ) {
3270 $this->mStripState
->nowiki
->setPair( $marker, $output );
3272 $this->mStripState
->general
->setPair( $marker, $output );
3278 * Increment an include size counter
3280 * @param string $type The type of expansion
3281 * @param integer $size The size of the text
3282 * @return boolean False if this inclusion would take it over the maximum, true otherwise
3284 function incrementIncludeSize( $type, $size ) {
3285 if ( $this->mIncludeSizes
[$type] +
$size > $this->mOptions
->getMaxIncludeSize( $type ) ) {
3288 $this->mIncludeSizes
[$type] +
= $size;
3294 * Increment the expensive function count
3296 * @return boolean False if the limit has been exceeded
3298 function incrementExpensiveFunctionCount() {
3299 global $wgExpensiveParserFunctionLimit;
3300 $this->mExpensiveFunctionCount++
;
3301 if($this->mExpensiveFunctionCount
<= $wgExpensiveParserFunctionLimit) {
3308 * Strip double-underscore items like __NOGALLERY__ and __NOTOC__
3309 * Fills $this->mDoubleUnderscores, returns the modified text
3311 function doDoubleUnderscore( $text ) {
3312 wfProfileIn( __METHOD__
);
3313 // The position of __TOC__ needs to be recorded
3314 $mw = MagicWord
::get( 'toc' );
3315 if( $mw->match( $text ) ) {
3316 $this->mShowToc
= true;
3317 $this->mForceTocPosition
= true;
3319 // Set a placeholder. At the end we'll fill it in with the TOC.
3320 $text = $mw->replace( '<!--MWTOC-->', $text, 1 );
3322 // Only keep the first one.
3323 $text = $mw->replace( '', $text );
3326 // Now match and remove the rest of them
3327 $mwa = MagicWord
::getDoubleUnderscoreArray();
3328 $this->mDoubleUnderscores
= $mwa->matchAndRemove( $text );
3330 if ( isset( $this->mDoubleUnderscores
['nogallery'] ) ) {
3331 $this->mOutput
->mNoGallery
= true;
3333 if ( isset( $this->mDoubleUnderscores
['notoc'] ) && !$this->mForceTocPosition
) {
3334 $this->mShowToc
= false;
3336 if ( isset( $this->mDoubleUnderscores
['hiddencat'] ) && $this->mTitle
->getNamespace() == NS_CATEGORY
) {
3337 $this->mOutput
->setProperty( 'hiddencat', 'y' );
3338 $this->addTrackingCategory( 'hidden-category-category' );
3340 # (bug 8068) Allow control over whether robots index a page.
3342 # FIXME (bug 14899): __INDEX__ always overrides __NOINDEX__ here! This
3343 # is not desirable, the last one on the page should win.
3344 if( isset( $this->mDoubleUnderscores
['noindex'] ) && $this->mTitle
->canUseNoindex() ) {
3345 $this->mOutput
->setIndexPolicy( 'noindex' );
3346 $this->addTrackingCategory( 'noindex-category' );
3348 if( isset( $this->mDoubleUnderscores
['index'] ) && $this->mTitle
->canUseNoindex() ){
3349 $this->mOutput
->setIndexPolicy( 'index' );
3350 $this->addTrackingCategory( 'index-category' );
3352 wfProfileOut( __METHOD__
);
3357 * Add a tracking category, getting the title from a system message,
3358 * or print a debug message if the title is invalid.
3359 * @param $msg String message key
3360 * @return Bool whether the addition was successful
3362 protected function addTrackingCategory( $msg ){
3363 $containerCategory = Title
::makeTitleSafe( NS_CATEGORY
, wfMsgForContent( $msg ) );
3364 if ( $containerCategory ) {
3365 $this->mOutput
->addCategory( $containerCategory->getDBkey(), $this->getDefaultSort() );
3368 wfDebug( __METHOD__
.": [[MediaWiki:$msg]] is not a valid title!\n" );
3374 * This function accomplishes several tasks:
3375 * 1) Auto-number headings if that option is enabled
3376 * 2) Add an [edit] link to sections for users who have enabled the option and can edit the page
3377 * 3) Add a Table of contents on the top for users who have enabled the option
3378 * 4) Auto-anchor headings
3380 * It loops through all headlines, collects the necessary data, then splits up the
3381 * string and re-inserts the newly formatted headlines.
3383 * @param string $text
3384 * @param string $origText Original, untouched wikitext
3385 * @param boolean $isMain
3388 function formatHeadings( $text, $origText, $isMain=true ) {
3389 global $wgMaxTocLevel, $wgContLang, $wgEnforceHtmlIds;
3391 $doNumberHeadings = $this->mOptions
->getNumberHeadings();
3392 $showEditLink = $this->mOptions
->getEditSection();
3394 // Do not call quickUserCan unless necessary
3395 if( $showEditLink && !$this->mTitle
->quickUserCan( 'edit' ) ) {
3399 # Inhibit editsection links if requested in the page
3400 if ( isset( $this->mDoubleUnderscores
['noeditsection'] ) ||
$this->mOptions
->getIsPrintable() ) {
3404 # Get all headlines for numbering them and adding funky stuff like [edit]
3405 # links - this is for later, but we need the number of headlines right now
3407 $numMatches = preg_match_all( '/<H(?P<level>[1-6])(?P<attrib>.*?'.'>)(?P<header>.*?)<\/H[1-6] *>/i', $text, $matches );
3409 # if there are fewer than 4 headlines in the article, do not show TOC
3410 # unless it's been explicitly enabled.
3411 $enoughToc = $this->mShowToc
&&
3412 (($numMatches >= 4) ||
$this->mForceTocPosition
);
3414 # Allow user to stipulate that a page should have a "new section"
3415 # link added via __NEWSECTIONLINK__
3416 if ( isset( $this->mDoubleUnderscores
['newsectionlink'] ) ) {
3417 $this->mOutput
->setNewSection( true );
3420 # Allow user to remove the "new section"
3421 # link via __NONEWSECTIONLINK__
3422 if ( isset( $this->mDoubleUnderscores
['nonewsectionlink'] ) ) {
3423 $this->mOutput
->hideNewSection( true );
3426 # if the string __FORCETOC__ (not case-sensitive) occurs in the HTML,
3427 # override above conditions and always show TOC above first header
3428 if ( isset( $this->mDoubleUnderscores
['forcetoc'] ) ) {
3429 $this->mShowToc
= true;
3433 # We need this to perform operations on the HTML
3434 $sk = $this->mOptions
->getSkin();
3440 # Ugh .. the TOC should have neat indentation levels which can be
3441 # passed to the skin functions. These are determined here
3445 $sublevelCount = array();
3446 $levelCount = array();
3452 $markerRegex = "{$this->mUniqPrefix}-h-(\d+)-" . self
::MARKER_SUFFIX
;
3453 $baseTitleText = $this->mTitle
->getPrefixedDBkey();
3454 $oldType = $this->mOutputType
;
3455 $this->setOutputType( self
::OT_WIKI
);
3456 $frame = $this->getPreprocessor()->newFrame();
3457 $root = $this->preprocessToDom( $origText );
3458 $node = $root->getFirstChild();
3462 foreach( $matches[3] as $headline ) {
3463 $isTemplate = false;
3465 $sectionIndex = false;
3467 $markerMatches = array();
3468 if (preg_match("/^$markerRegex/", $headline, $markerMatches)) {
3469 $serial = $markerMatches[1];
3470 list( $titleText, $sectionIndex ) = $this->mHeadings
[$serial];
3471 $isTemplate = ($titleText != $baseTitleText);
3472 $headline = preg_replace("/^$markerRegex/", "", $headline);
3476 $prevlevel = $level;
3477 $prevtoclevel = $toclevel;
3479 $level = $matches[1][$headlineCount];
3481 if ( $level > $prevlevel ) {
3482 # Increase TOC level
3484 $sublevelCount[$toclevel] = 0;
3485 if( $toclevel<$wgMaxTocLevel ) {
3486 $prevtoclevel = $toclevel;
3487 $toc .= $sk->tocIndent();
3491 elseif ( $level < $prevlevel && $toclevel > 1 ) {
3492 # Decrease TOC level, find level to jump to
3494 for ($i = $toclevel; $i > 0; $i--) {
3495 if ( $levelCount[$i] == $level ) {
3496 # Found last matching level
3500 elseif ( $levelCount[$i] < $level ) {
3501 # Found first matching level below current level
3506 if( $i == 0 ) $toclevel = 1;
3507 if( $toclevel<$wgMaxTocLevel ) {
3508 if($prevtoclevel < $wgMaxTocLevel) {
3509 # Unindent only if the previous toc level was shown :p
3510 $toc .= $sk->tocUnindent( $prevtoclevel - $toclevel );
3511 $prevtoclevel = $toclevel;
3513 $toc .= $sk->tocLineEnd();
3518 # No change in level, end TOC line
3519 if( $toclevel<$wgMaxTocLevel ) {
3520 $toc .= $sk->tocLineEnd();
3524 $levelCount[$toclevel] = $level;
3526 # count number of headlines for each level
3527 @$sublevelCount[$toclevel]++
;
3529 for( $i = 1; $i <= $toclevel; $i++
) {
3530 if( !empty( $sublevelCount[$i] ) ) {
3534 $numbering .= $wgContLang->formatNum( $sublevelCount[$i] );
3539 # The safe header is a version of the header text safe to use for links
3540 # Avoid insertion of weird stuff like <math> by expanding the relevant sections
3541 $safeHeadline = $this->mStripState
->unstripBoth( $headline );
3543 # Remove link placeholders by the link text.
3544 # <!--LINK number-->
3546 # link text with suffix
3547 $safeHeadline = $this->replaceLinkHoldersText( $safeHeadline );
3549 # Strip out HTML (other than plain <sup> and <sub>: bug 8393)
3550 $tocline = preg_replace(
3551 array( '#<(?!/?(sup|sub)).*?'.'>#', '#<(/?(sup|sub)).*?'.'>#' ),
3555 $tocline = trim( $tocline );
3557 # For the anchor, strip out HTML-y stuff period
3558 $safeHeadline = preg_replace( '/<.*?'.'>/', '', $safeHeadline );
3559 $safeHeadline = preg_replace( '/[ _]+/', ' ', $safeHeadline );
3560 $safeHeadline = trim( $safeHeadline );
3562 # Save headline for section edit hint before it's escaped
3563 $headlineHint = $safeHeadline;
3565 if ( $wgEnforceHtmlIds ) {
3566 $legacyHeadline = false;
3567 $safeHeadline = Sanitizer
::escapeId( $safeHeadline,
3570 # For reverse compatibility, provide an id that's
3571 # HTML4-compatible, like we used to.
3573 # It may be worth noting, academically, that it's possible for
3574 # the legacy anchor to conflict with a non-legacy headline
3575 # anchor on the page. In this case likely the "correct" thing
3576 # would be to either drop the legacy anchors or make sure
3577 # they're numbered first. However, this would require people
3578 # to type in section names like "abc_.D7.93.D7.90.D7.A4"
3579 # manually, so let's not bother worrying about it.
3580 $legacyHeadline = Sanitizer
::escapeId( $safeHeadline,
3582 $safeHeadline = Sanitizer
::escapeId( $safeHeadline, 'xml' );
3584 if ( $legacyHeadline == $safeHeadline ) {
3585 # No reason to have both (in fact, we can't)
3586 $legacyHeadline = false;
3587 } elseif ( $legacyHeadline != Sanitizer
::escapeId(
3588 $legacyHeadline, 'xml' ) ) {
3589 # The legacy id is invalid XML. We used to allow this, but
3590 # there's no reason to do so anymore. Backward
3591 # compatibility will fail slightly in this case, but it's
3593 $legacyHeadline = false;
3597 # HTML names must be case-insensitively unique (bug 10721). FIXME:
3598 # Does this apply to Unicode characters? Because we aren't
3599 # handling those here.
3600 $arrayKey = strtolower( $safeHeadline );
3601 if ( $legacyHeadline === false ) {
3602 $legacyArrayKey = false;
3604 $legacyArrayKey = strtolower( $legacyHeadline );
3607 # count how many in assoc. array so we can track dupes in anchors
3608 if ( isset( $refers[$arrayKey] ) ) {
3609 $refers[$arrayKey]++
;
3611 $refers[$arrayKey] = 1;
3613 if ( isset( $refers[$legacyArrayKey] ) ) {
3614 $refers[$legacyArrayKey]++
;
3616 $refers[$legacyArrayKey] = 1;
3619 # Don't number the heading if it is the only one (looks silly)
3620 if( $doNumberHeadings && count( $matches[3] ) > 1) {
3621 # the two are different if the line contains a link
3622 $headline=$numbering . ' ' . $headline;
3625 # Create the anchor for linking from the TOC to the section
3626 $anchor = $safeHeadline;
3627 $legacyAnchor = $legacyHeadline;
3628 if ( $refers[$arrayKey] > 1 ) {
3629 $anchor .= '_' . $refers[$arrayKey];
3631 if ( $legacyHeadline !== false && $refers[$legacyArrayKey] > 1 ) {
3632 $legacyAnchor .= '_' . $refers[$legacyArrayKey];
3634 if( $enoughToc && ( !isset($wgMaxTocLevel) ||
$toclevel<$wgMaxTocLevel ) ) {
3635 $toc .= $sk->tocLine($anchor, $tocline,
3636 $numbering, $toclevel, ($isTemplate ?
false : $sectionIndex));
3639 # Add the section to the section tree
3640 # Find the DOM node for this header
3641 while ( $node && !$isTemplate ) {
3642 if ( $node->getName() === 'h' ) {
3643 $bits = $node->splitHeading();
3644 if ( $bits['i'] == $sectionIndex )
3647 $byteOffset +
= mb_strlen( $this->mStripState
->unstripBoth(
3648 $frame->expand( $node, PPFrame
::RECOVER_ORIG
) ) );
3649 $node = $node->getNextSibling();
3652 'toclevel' => $toclevel,
3655 'number' => $numbering,
3656 'index' => ($isTemplate ?
'T-' : '' ) . $sectionIndex,
3657 'fromtitle' => $titleText,
3658 'byteoffset' => ( $isTemplate ?
null : $byteOffset ),
3659 'anchor' => $anchor,
3662 # give headline the correct <h#> tag
3663 if( $showEditLink && $sectionIndex !== false ) {
3665 # Put a T flag in the section identifier, to indicate to extractSections()
3666 # that sections inside <includeonly> should be counted.
3667 $editlink = $sk->doEditSectionLink(Title
::newFromText( $titleText ), "T-$sectionIndex");
3669 $editlink = $sk->doEditSectionLink($this->mTitle
, $sectionIndex, $headlineHint);
3674 $head[$headlineCount] = $sk->makeHeadline( $level,
3675 $matches['attrib'][$headlineCount], $anchor, $headline,
3676 $editlink, $legacyAnchor );
3681 $this->setOutputType( $oldType );
3683 # Never ever show TOC if no headers
3684 if( $numVisible < 1 ) {
3689 if( $prevtoclevel > 0 && $prevtoclevel < $wgMaxTocLevel ) {
3690 $toc .= $sk->tocUnindent( $prevtoclevel - 1 );
3692 $toc = $sk->tocList( $toc );
3693 $this->mOutput
->setTOCHTML( $toc );
3697 $this->mOutput
->setSections( $tocraw );
3700 # split up and insert constructed headlines
3702 $blocks = preg_split( '/<H[1-6].*?' . '>.*?<\/H[1-6]>/i', $text );
3705 foreach( $blocks as $block ) {
3706 if( $showEditLink && $headlineCount > 0 && $i == 0 && $block !== "\n" ) {
3707 # This is the [edit] link that appears for the top block of text when
3708 # section editing is enabled
3710 # Disabled because it broke block formatting
3711 # For example, a bullet point in the top line
3712 # $full .= $sk->editSectionLink(0);
3715 if( $enoughToc && !$i && $isMain && !$this->mForceTocPosition
) {
3716 # Top anchor now in skin
3720 if( !empty( $head[$i] ) ) {
3725 if( $this->mForceTocPosition
) {
3726 return str_replace( '<!--MWTOC-->', $toc, $full );
3733 * Merge $tree2 into $tree1 by replacing the section with index
3734 * $section in $tree1 and its descendants with the sections in $tree2.
3735 * Note that in the returned section tree, only the 'index' and
3736 * 'byteoffset' fields are guaranteed to be correct.
3737 * @param $tree1 array Section tree from ParserOutput::getSectons()
3738 * @param $tree2 array Section tree
3739 * @param $section int Section index
3740 * @param $title Title Title both section trees come from
3741 * @param $len2 int Length of the original wikitext for $tree2
3742 * @return array Merged section tree
3744 public static function mergeSectionTrees( $tree1, $tree2, $section, $title, $len2 ) {
3747 $targetLevel = false;
3751 $numbering = array( 0 );
3752 $titletext = $title->getPrefixedDBkey();
3753 foreach ( $tree1 as $s ) {
3754 if ( $targetLevel !== false ) {
3755 if ( $s['level'] <= $targetLevel )
3756 // We've skipped enough
3757 $targetLevel = false;
3761 if ( $s['index'] != $section ||
3762 $s['fromtitle'] != $titletext ) {
3763 self
::incrementNumbering( $numbering,
3764 $s['toclevel'], $lastLevel );
3766 // Rewrite index, byteoffset and number
3767 if ( $s['fromtitle'] == $titletext ) {
3768 $s['index'] = $nextIndex++
;
3770 $s['byteoffset'] +
= $len2;
3772 $s['number'] = implode( '.', array_map(
3773 array( $wgContLang, 'formatnum' ),
3775 $lastLevel = $s['toclevel'];
3778 // We're at $section
3779 // Insert sections from $tree2 here
3780 foreach ( $tree2 as $s2 ) {
3781 // Rewrite the fields in $s2
3782 // before inserting it
3783 $s2['toclevel'] +
= $s['toclevel'] - 1;
3784 $s2['level'] +
= $s['level'] - 1;
3785 $s2['index'] = $nextIndex++
;
3786 $s2['byteoffset'] +
= $s['byteoffset'];
3788 self
::incrementNumbering( $numbering,
3789 $s2['toclevel'], $lastLevel );
3790 $s2['number'] = implode( '.', array_map(
3791 array( $wgContLang, 'formatnum' ),
3793 $lastLevel = $s2['toclevel'];
3796 // Skip all descendants of $section in $tree1
3797 $targetLevel = $s['level'];
3805 * Increment a section number. Helper function for mergeSectionTrees()
3806 * @param $number array Array representing a section number
3807 * @param $level int Current TOC level (depth)
3808 * @param $lastLevel int Level of previous TOC entry
3810 private static function incrementNumbering( &$number, $level, $lastLevel ) {
3811 if ( $level > $lastLevel )
3812 $number[$level - 1] = 1;
3813 else if ( $level < $lastLevel ) {
3814 foreach ( $number as $key => $unused )
3815 if ( $key >= $level )
3816 unset( $number[$key] );
3817 $number[$level - 1]++
;
3819 $number[$level - 1]++
;
3823 * Transform wiki markup when saving a page by doing \r\n -> \n
3824 * conversion, substitting signatures, {{subst:}} templates, etc.
3826 * @param string $text the text to transform
3827 * @param Title &$title the Title object for the current article
3828 * @param User $user the User object describing the current user
3829 * @param ParserOptions $options parsing options
3830 * @param bool $clearState whether to clear the parser state first
3831 * @return string the altered wiki markup
3834 function preSaveTransform( $text, Title
$title, $user, $options, $clearState = true ) {
3835 $this->mOptions
= $options;
3836 $this->setTitle( $title );
3837 $this->setOutputType( self
::OT_WIKI
);
3839 if ( $clearState ) {
3840 $this->clearState();
3846 $text = str_replace( array_keys( $pairs ), array_values( $pairs ), $text );
3847 $text = $this->pstPass2( $text, $user );
3848 $text = $this->mStripState
->unstripBoth( $text );
3853 * Pre-save transform helper function
3856 function pstPass2( $text, $user ) {
3857 global $wgContLang, $wgLocaltimezone;
3859 /* Note: This is the timestamp saved as hardcoded wikitext to
3860 * the database, we use $wgContLang here in order to give
3861 * everyone the same signature and use the default one rather
3862 * than the one selected in each user's preferences.
3864 * (see also bug 12815)
3866 $ts = $this->mOptions
->getTimestamp();
3867 $tz = wfMsgForContent( 'timezone-utc' );
3868 if ( isset( $wgLocaltimezone ) ) {
3869 $unixts = wfTimestamp( TS_UNIX
, $ts );
3870 $oldtz = getenv( 'TZ' );
3871 putenv( 'TZ='.$wgLocaltimezone );
3872 $ts = date( 'YmdHis', $unixts );
3873 $tz = date( 'T', $unixts ); # might vary on DST changeover!
3875 /* Allow translation of timezones trough wiki. date() can return
3876 * whatever crap the system uses, localised or not, so we cannot
3877 * ship premade translations.
3879 $key = 'timezone-' . strtolower( trim( $tz ) );
3880 $value = wfMsgForContent( $key );
3881 if ( !wfEmptyMsg( $key, $value ) ) $tz = $value;
3883 putenv( 'TZ='.$oldtz );
3886 $d = $wgContLang->timeanddate( $ts, false, false ) . " ($tz)";
3888 # Variable replacement
3889 # Because mOutputType is OT_WIKI, this will only process {{subst:xxx}} type tags
3890 $text = $this->replaceVariables( $text );
3893 $sigText = $this->getUserSig( $user );
3894 $text = strtr( $text, array(
3896 '~~~~' => "$sigText $d",
3900 # Context links: [[|name]] and [[name (context)|]]
3902 global $wgLegalTitleChars;
3903 $tc = "[$wgLegalTitleChars]";
3904 $nc = '[ _0-9A-Za-z\x80-\xff-]'; # Namespaces can use non-ascii!
3906 $p1 = "/\[\[(:?$nc+:|:|)($tc+?)( \\($tc+\\))\\|]]/"; # [[ns:page (context)|]]
3907 $p4 = "/\[\[(:?$nc+:|:|)($tc+?)(($tc+))\\|]]/"; # [[ns:page(context)|]]
3908 $p3 = "/\[\[(:?$nc+:|:|)($tc+?)( \\($tc+\\)|)(, $tc+|)\\|]]/"; # [[ns:page (context), context|]]
3909 $p2 = "/\[\[\\|($tc+)]]/"; # [[|page]]
3911 # try $p1 first, to turn "[[A, B (C)|]]" into "[[A, B (C)|A, B]]"
3912 $text = preg_replace( $p1, '[[\\1\\2\\3|\\2]]', $text );
3913 $text = preg_replace( $p4, '[[\\1\\2\\3|\\2]]', $text );
3914 $text = preg_replace( $p3, '[[\\1\\2\\3\\4|\\2]]', $text );
3916 $t = $this->mTitle
->getText();
3918 if ( preg_match( "/^($nc+:|)$tc+?( \\($tc+\\))$/", $t, $m ) ) {
3919 $text = preg_replace( $p2, "[[$m[1]\\1$m[2]|\\1]]", $text );
3920 } elseif ( preg_match( "/^($nc+:|)$tc+?(, $tc+|)$/", $t, $m ) && '' != "$m[1]$m[2]" ) {
3921 $text = preg_replace( $p2, "[[$m[1]\\1$m[2]|\\1]]", $text );
3923 # if there's no context, don't bother duplicating the title
3924 $text = preg_replace( $p2, '[[\\1]]', $text );
3927 # Trim trailing whitespace
3928 $text = rtrim( $text );
3934 * Fetch the user's signature text, if any, and normalize to
3935 * validated, ready-to-insert wikitext.
3936 * If you have pre-fetched the nickname or the fancySig option, you can
3937 * specify them here to save a database query.
3942 function getUserSig( &$user, $nickname = false, $fancySig = null ) {
3943 global $wgMaxSigChars;
3945 $username = $user->getName();
3947 // If not given, retrieve from the user object.
3948 if ( $nickname === false )
3949 $nickname = $user->getOption( 'nickname' );
3951 if ( is_null( $fancySig ) )
3952 $fancySig = $user->getBoolOption( 'fancysig' );
3954 $nickname = $nickname == null ?
$username : $nickname;
3956 if( mb_strlen( $nickname ) > $wgMaxSigChars ) {
3957 $nickname = $username;
3958 wfDebug( __METHOD__
. ": $username has overlong signature.\n" );
3959 } elseif( $fancySig !== false ) {
3960 # Sig. might contain markup; validate this
3961 if( $this->validateSig( $nickname ) !== false ) {
3962 # Validated; clean up (if needed) and return it
3963 return $this->cleanSig( $nickname, true );
3965 # Failed to validate; fall back to the default
3966 $nickname = $username;
3967 wfDebug( __METHOD__
.": $username has bad XML tags in signature.\n" );
3971 // Make sure nickname doesnt get a sig in a sig
3972 $nickname = $this->cleanSigInSig( $nickname );
3974 # If we're still here, make it a link to the user page
3975 $userText = wfEscapeWikiText( $username );
3976 $nickText = wfEscapeWikiText( $nickname );
3977 if ( $user->isAnon() ) {
3978 return wfMsgExt( 'signature-anon', array( 'content', 'parsemag' ), $userText, $nickText );
3980 return wfMsgExt( 'signature', array( 'content', 'parsemag' ), $userText, $nickText );
3985 * Check that the user's signature contains no bad XML
3987 * @param string $text
3988 * @return mixed An expanded string, or false if invalid.
3990 function validateSig( $text ) {
3991 return( Xml
::isWellFormedXmlFragment( $text ) ?
$text : false );
3995 * Clean up signature text
3997 * 1) Strip ~~~, ~~~~ and ~~~~~ out of signatures @see cleanSigInSig
3998 * 2) Substitute all transclusions
4000 * @param string $text
4001 * @param $parsing Whether we're cleaning (preferences save) or parsing
4002 * @return string Signature text
4004 function cleanSig( $text, $parsing = false ) {
4007 $this->clearState();
4008 $this->setTitle( $wgTitle );
4009 $this->mOptions
= new ParserOptions
;
4010 $this->setOutputType
= self
::OT_PREPROCESS
;
4013 # Option to disable this feature
4014 if ( !$this->mOptions
->getCleanSignatures() ) {
4018 # FIXME: regex doesn't respect extension tags or nowiki
4019 # => Move this logic to braceSubstitution()
4020 $substWord = MagicWord
::get( 'subst' );
4021 $substRegex = '/\{\{(?!(?:' . $substWord->getBaseRegex() . '))/x' . $substWord->getRegexCase();
4022 $substText = '{{' . $substWord->getSynonym( 0 );
4024 $text = preg_replace( $substRegex, $substText, $text );
4025 $text = $this->cleanSigInSig( $text );
4026 $dom = $this->preprocessToDom( $text );
4027 $frame = $this->getPreprocessor()->newFrame();
4028 $text = $frame->expand( $dom );
4031 $text = $this->mStripState
->unstripBoth( $text );
4038 * Strip ~~~, ~~~~ and ~~~~~ out of signatures
4039 * @param string $text
4040 * @return string Signature text with /~{3,5}/ removed
4042 function cleanSigInSig( $text ) {
4043 $text = preg_replace( '/~{3,5}/', '', $text );
4048 * Set up some variables which are usually set up in parse()
4049 * so that an external function can call some class members with confidence
4052 function startExternalParse( &$title, $options, $outputType, $clearState = true ) {
4053 $this->setTitle( $title );
4054 $this->mOptions
= $options;
4055 $this->setOutputType( $outputType );
4056 if ( $clearState ) {
4057 $this->clearState();
4062 * Wrapper for preprocess()
4064 * @param string $text the text to preprocess
4065 * @param ParserOptions $options options
4069 function transformMsg( $text, $options ) {
4071 static $executing = false;
4073 # Guard against infinite recursion
4079 wfProfileIn(__METHOD__
);
4080 $text = $this->preprocess( $text, $wgTitle, $options );
4083 wfProfileOut(__METHOD__
);
4088 * Create an HTML-style tag, e.g. <yourtag>special text</yourtag>
4089 * The callback should have the following form:
4090 * function myParserHook( $text, $params, &$parser ) { ... }
4092 * Transform and return $text. Use $parser for any required context, e.g. use
4093 * $parser->getTitle() and $parser->getOptions() not $wgTitle or $wgOut->mParserOptions
4097 * @param mixed $tag The tag to use, e.g. 'hook' for <hook>
4098 * @param mixed $callback The callback function (and object) to use for the tag
4100 * @return The old value of the mTagHooks array associated with the hook
4102 function setHook( $tag, $callback ) {
4103 $tag = strtolower( $tag );
4104 $oldVal = isset( $this->mTagHooks
[$tag] ) ?
$this->mTagHooks
[$tag] : null;
4105 $this->mTagHooks
[$tag] = $callback;
4106 if( !in_array( $tag, $this->mStripList
) ) {
4107 $this->mStripList
[] = $tag;
4113 function setTransparentTagHook( $tag, $callback ) {
4114 $tag = strtolower( $tag );
4115 $oldVal = isset( $this->mTransparentTagHooks
[$tag] ) ?
$this->mTransparentTagHooks
[$tag] : null;
4116 $this->mTransparentTagHooks
[$tag] = $callback;
4122 * Remove all tag hooks
4124 function clearTagHooks() {
4125 $this->mTagHooks
= array();
4126 $this->mStripList
= $this->mDefaultStripList
;
4130 * Create a function, e.g. {{sum:1|2|3}}
4131 * The callback function should have the form:
4132 * function myParserFunction( &$parser, $arg1, $arg2, $arg3 ) { ... }
4134 * Or with SFH_OBJECT_ARGS:
4135 * function myParserFunction( $parser, $frame, $args ) { ... }
4137 * The callback may either return the text result of the function, or an array with the text
4138 * in element 0, and a number of flags in the other elements. The names of the flags are
4139 * specified in the keys. Valid flags are:
4140 * found The text returned is valid, stop processing the template. This
4142 * nowiki Wiki markup in the return value should be escaped
4143 * isHTML The returned text is HTML, armour it against wikitext transformation
4147 * @param string $id The magic word ID
4148 * @param mixed $callback The callback function (and object) to use
4149 * @param integer $flags a combination of the following flags:
4150 * SFH_NO_HASH No leading hash, i.e. {{plural:...}} instead of {{#if:...}}
4152 * SFH_OBJECT_ARGS Pass the template arguments as PPNode objects instead of text. This
4153 * allows for conditional expansion of the parse tree, allowing you to eliminate dead
4154 * branches and thus speed up parsing. It is also possible to analyse the parse tree of
4155 * the arguments, and to control the way they are expanded.
4157 * The $frame parameter is a PPFrame. This can be used to produce expanded text from the
4158 * arguments, for instance:
4159 * $text = isset( $args[0] ) ? $frame->expand( $args[0] ) : '';
4161 * For technical reasons, $args[0] is pre-expanded and will be a string. This may change in
4162 * future versions. Please call $frame->expand() on it anyway so that your code keeps
4163 * working if/when this is changed.
4165 * If you want whitespace to be trimmed from $args, you need to do it yourself, post-
4168 * Please read the documentation in includes/parser/Preprocessor.php for more information
4169 * about the methods available in PPFrame and PPNode.
4171 * @return The old callback function for this name, if any
4173 function setFunctionHook( $id, $callback, $flags = 0 ) {
4176 $oldVal = isset( $this->mFunctionHooks
[$id] ) ?
$this->mFunctionHooks
[$id][0] : null;
4177 $this->mFunctionHooks
[$id] = array( $callback, $flags );
4179 # Add to function cache
4180 $mw = MagicWord
::get( $id );
4182 throw new MWException( __METHOD__
.'() expecting a magic word identifier.' );
4184 $synonyms = $mw->getSynonyms();
4185 $sensitive = intval( $mw->isCaseSensitive() );
4187 foreach ( $synonyms as $syn ) {
4189 if ( !$sensitive ) {
4190 $syn = $wgContLang->lc( $syn );
4193 if ( !( $flags & SFH_NO_HASH
) ) {
4196 # Remove trailing colon
4197 if ( substr( $syn, -1, 1 ) === ':' ) {
4198 $syn = substr( $syn, 0, -1 );
4200 $this->mFunctionSynonyms
[$sensitive][$syn] = $id;
4206 * Get all registered function hook identifiers
4210 function getFunctionHooks() {
4211 return array_keys( $this->mFunctionHooks
);
4215 * Create a tag function, e.g. <test>some stuff</test>.
4216 * Unlike tag hooks, tag functions are parsed at preprocessor level.
4217 * Unlike parser functions, their content is not preprocessed.
4219 function setFunctionTagHook( $tag, $callback, $flags ) {
4220 $tag = strtolower( $tag );
4221 $old = isset( $this->mFunctionTagHooks
[$tag] ) ?
4222 $this->mFunctionTagHooks
[$tag] : null;
4223 $this->mFunctionTagHooks
[$tag] = array( $callback, $flags );
4225 if( !in_array( $tag, $this->mStripList
) ) {
4226 $this->mStripList
[] = $tag;
4233 * FIXME: update documentation. makeLinkObj() is deprecated.
4234 * Replace <!--LINK--> link placeholders with actual links, in the buffer
4235 * Placeholders created in Skin::makeLinkObj()
4236 * Returns an array of link CSS classes, indexed by PDBK.
4238 function replaceLinkHolders( &$text, $options = 0 ) {
4239 return $this->mLinkHolders
->replace( $text );
4243 * Replace <!--LINK--> link placeholders with plain text of links
4244 * (not HTML-formatted).
4245 * @param string $text
4248 function replaceLinkHoldersText( $text ) {
4249 return $this->mLinkHolders
->replaceText( $text );
4253 * Tag hook handler for 'pre'.
4255 function renderPreTag( $text, $attribs ) {
4256 // Backwards-compatibility hack
4257 $content = StringUtils
::delimiterReplace( '<nowiki>', '</nowiki>', '$1', $text, 'i' );
4259 $attribs = Sanitizer
::validateTagAttributes( $attribs, 'pre' );
4260 return Xml
::openElement( 'pre', $attribs ) .
4261 Xml
::escapeTagsOnly( $content ) .
4266 * Renders an image gallery from a text with one line per image.
4267 * text labels may be given by using |-style alternative text. E.g.
4268 * Image:one.jpg|The number "1"
4269 * Image:tree.jpg|A tree
4270 * given as text will return the HTML of a gallery with two images,
4271 * labeled 'The number "1"' and
4274 function renderImageGallery( $text, $params ) {
4275 $ig = new ImageGallery();
4276 $ig->setContextTitle( $this->mTitle
);
4277 $ig->setShowBytes( false );
4278 $ig->setShowFilename( false );
4279 $ig->setParser( $this );
4280 $ig->setHideBadImages();
4281 $ig->setAttributes( Sanitizer
::validateTagAttributes( $params, 'table' ) );
4282 $ig->useSkin( $this->mOptions
->getSkin() );
4283 $ig->mRevisionId
= $this->mRevisionId
;
4285 if( isset( $params['caption'] ) ) {
4286 $caption = $params['caption'];
4287 $caption = htmlspecialchars( $caption );
4288 $caption = $this->replaceInternalLinks( $caption );
4289 $ig->setCaptionHtml( $caption );
4291 if( isset( $params['perrow'] ) ) {
4292 $ig->setPerRow( $params['perrow'] );
4294 if( isset( $params['widths'] ) ) {
4295 $ig->setWidths( $params['widths'] );
4297 if( isset( $params['heights'] ) ) {
4298 $ig->setHeights( $params['heights'] );
4301 wfRunHooks( 'BeforeParserrenderImageGallery', array( &$this, &$ig ) );
4303 $lines = StringUtils
::explode( "\n", $text );
4304 foreach ( $lines as $line ) {
4305 # match lines like these:
4306 # Image:someimage.jpg|This is some image
4308 preg_match( "/^([^|]+)(\\|(.*))?$/", $line, $matches );
4310 if ( count( $matches ) == 0 ) {
4314 if ( strpos( $matches[0], '%' ) !== false )
4315 $matches[1] = urldecode( $matches[1] );
4316 $tp = Title
::newFromText( $matches[1]/*, NS_FILE*/ );
4318 if( is_null( $nt ) ) {
4319 # Bogus title. Ignore these so we don't bomb out later.
4322 if ( isset( $matches[3] ) ) {
4323 $label = $matches[3];
4328 $html = $this->recursiveTagParse( trim( $label ) );
4330 $ig->add( $nt, $html );
4332 # Only add real images (bug #5586)
4333 if ( $nt->getNamespace() == NS_FILE
) {
4334 $this->mOutput
->addImage( $nt->getDBkey() );
4337 return $ig->toHTML();
4340 function getImageParams( $handler ) {
4342 $handlerClass = get_class( $handler );
4346 if ( !isset( $this->mImageParams
[$handlerClass] ) ) {
4347 // Initialise static lists
4348 static $internalParamNames = array(
4349 'horizAlign' => array( 'left', 'right', 'center', 'none' ),
4350 'vertAlign' => array( 'baseline', 'sub', 'super', 'top', 'text-top', 'middle',
4351 'bottom', 'text-bottom' ),
4352 'frame' => array( 'thumbnail', 'manualthumb', 'framed', 'frameless',
4353 'upright', 'border', 'link', 'alt' ),
4355 static $internalParamMap;
4356 if ( !$internalParamMap ) {
4357 $internalParamMap = array();
4358 foreach ( $internalParamNames as $type => $names ) {
4359 foreach ( $names as $name ) {
4360 $magicName = str_replace( '-', '_', "img_$name" );
4361 $internalParamMap[$magicName] = array( $type, $name );
4366 // Add handler params
4367 $paramMap = $internalParamMap;
4369 $handlerParamMap = $handler->getParamMap();
4370 foreach ( $handlerParamMap as $magic => $paramName ) {
4371 $paramMap[$magic] = array( 'handler', $paramName );
4374 $this->mImageParams
[$handlerClass] = $paramMap;
4375 $this->mImageParamsMagicArray
[$handlerClass] = new MagicWordArray( array_keys( $paramMap ) );
4377 return array( $this->mImageParams
[$handlerClass], $this->mImageParamsMagicArray
[$handlerClass] );
4381 * Parse image options text and use it to make an image
4382 * @param Title $title
4383 * @param string $options
4384 * @param LinkHolderArray $holders
4386 function makeImage( $title, $options, $holders = false ) {
4387 # Check if the options text is of the form "options|alt text"
4389 # * thumbnail make a thumbnail with enlarge-icon and caption, alignment depends on lang
4390 # * left no resizing, just left align. label is used for alt= only
4391 # * right same, but right aligned
4392 # * none same, but not aligned
4393 # * ___px scale to ___ pixels width, no aligning. e.g. use in taxobox
4394 # * center center the image
4395 # * frame Keep original image size, no magnify-button.
4396 # * framed Same as "frame"
4397 # * frameless like 'thumb' but without a frame. Keeps user preferences for width
4398 # * upright reduce width for upright images, rounded to full __0 px
4399 # * border draw a 1px border around the image
4400 # * alt Text for HTML alt attribute (defaults to empty)
4401 # * link Set the target of the image link. Can be external, interwiki, or local
4402 # vertical-align values (no % or length right now):
4412 $parts = StringUtils
::explode( "|", $options );
4413 $sk = $this->mOptions
->getSkin();
4415 # Give extensions a chance to select the file revision for us
4416 $skip = $time = $descQuery = false;
4417 wfRunHooks( 'BeforeParserMakeImageLinkObj', array( &$this, &$title, &$skip, &$time, &$descQuery ) );
4420 return $sk->link( $title );
4424 $imagename = $title->getDBkey();
4425 $file = wfFindFile( $title, array( 'time' => $time ) );
4427 $handler = $file ?
$file->getHandler() : false;
4429 list( $paramMap, $mwArray ) = $this->getImageParams( $handler );
4431 # Process the input parameters
4433 $params = array( 'frame' => array(), 'handler' => array(),
4434 'horizAlign' => array(), 'vertAlign' => array() );
4435 foreach( $parts as $part ) {
4436 $part = trim( $part );
4437 list( $magicName, $value ) = $mwArray->matchVariableStartToEnd( $part );
4439 if( isset( $paramMap[$magicName] ) ) {
4440 list( $type, $paramName ) = $paramMap[$magicName];
4442 // Special case; width and height come in one variable together
4443 if( $type === 'handler' && $paramName === 'width' ) {
4445 # (bug 13500) In both cases (width/height and width only),
4446 # permit trailing "px" for backward compatibility.
4447 if ( preg_match( '/^([0-9]*)x([0-9]*)\s*(?:px)?\s*$/', $value, $m ) ) {
4448 $width = intval( $m[1] );
4449 $height = intval( $m[2] );
4450 if ( $handler->validateParam( 'width', $width ) ) {
4451 $params[$type]['width'] = $width;
4454 if ( $handler->validateParam( 'height', $height ) ) {
4455 $params[$type]['height'] = $height;
4458 } elseif ( preg_match( '/^[0-9]*\s*(?:px)?\s*$/', $value ) ) {
4459 $width = intval( $value );
4460 if ( $handler->validateParam( 'width', $width ) ) {
4461 $params[$type]['width'] = $width;
4464 } // else no validation -- bug 13436
4466 if ( $type === 'handler' ) {
4467 # Validate handler parameter
4468 $validated = $handler->validateParam( $paramName, $value );
4470 # Validate internal parameters
4471 switch( $paramName ) {
4474 // @fixme - possibly check validity here for
4475 // manualthumb? downstream behavior seems odd with
4476 // missing manual thumbs.
4478 $value = $this->stripAltText( $value, $holders );
4481 $chars = self
::EXT_LINK_URL_CLASS
;
4482 $prots = $this->mUrlProtocols
;
4483 if ( $value === '' ) {
4484 $paramName = 'no-link';
4487 } elseif ( preg_match( "/^$prots/", $value ) ) {
4488 if ( preg_match( "/^($prots)$chars+$/", $value, $m ) ) {
4489 $paramName = 'link-url';
4490 $this->mOutput
->addExternalLink( $value );
4494 $linkTitle = Title
::newFromText( $value );
4496 $paramName = 'link-title';
4497 $value = $linkTitle;
4498 $this->mOutput
->addLink( $linkTitle );
4504 // Most other things appear to be empty or numeric...
4505 $validated = ( $value === false ||
is_numeric( trim( $value ) ) );
4510 $params[$type][$paramName] = $value;
4514 if ( !$validated ) {
4519 # Process alignment parameters
4520 if ( $params['horizAlign'] ) {
4521 $params['frame']['align'] = key( $params['horizAlign'] );
4523 if ( $params['vertAlign'] ) {
4524 $params['frame']['valign'] = key( $params['vertAlign'] );
4527 $params['frame']['caption'] = $caption;
4529 # Will the image be presented in a frame, with the caption below?
4530 $imageIsFramed = isset( $params['frame']['frame'] ) ||
4531 isset( $params['frame']['framed'] ) ||
4532 isset( $params['frame']['thumbnail'] ) ||
4533 isset( $params['frame']['manualthumb'] );
4535 # In the old days, [[Image:Foo|text...]] would set alt text. Later it
4536 # came to also set the caption, ordinary text after the image -- which
4537 # makes no sense, because that just repeats the text multiple times in
4538 # screen readers. It *also* came to set the title attribute.
4540 # Now that we have an alt attribute, we should not set the alt text to
4541 # equal the caption: that's worse than useless, it just repeats the
4542 # text. This is the framed/thumbnail case. If there's no caption, we
4543 # use the unnamed parameter for alt text as well, just for the time be-
4544 # ing, if the unnamed param is set and the alt param is not.
4546 # For the future, we need to figure out if we want to tweak this more,
4547 # e.g., introducing a title= parameter for the title; ignoring the un-
4548 # named parameter entirely for images without a caption; adding an ex-
4549 # plicit caption= parameter and preserving the old magic unnamed para-
4551 if ( $imageIsFramed ) { # Framed image
4552 if ( $caption === '' && !isset( $params['frame']['alt'] ) ) {
4553 # No caption or alt text, add the filename as the alt text so
4554 # that screen readers at least get some description of the image
4555 $params['frame']['alt'] = $title->getText();
4557 # Do not set $params['frame']['title'] because tooltips don't make sense
4559 } else { # Inline image
4560 if ( !isset( $params['frame']['alt'] ) ) {
4561 # No alt text, use the "caption" for the alt text
4562 if ( $caption !== '') {
4563 $params['frame']['alt'] = $this->stripAltText( $caption, $holders );
4565 # No caption, fall back to using the filename for the
4567 $params['frame']['alt'] = $title->getText();
4570 # Use the "caption" for the tooltip text
4571 $params['frame']['title'] = $this->stripAltText( $caption, $holders );
4574 wfRunHooks( 'ParserMakeImageParams', array( $title, $file, &$params ) );
4576 # Linker does the rest
4577 $ret = $sk->makeImageLink2( $title, $file, $params['frame'], $params['handler'], $time, $descQuery );
4579 # Give the handler a chance to modify the parser object
4581 $handler->parserTransformHook( $this, $file );
4587 protected function stripAltText( $caption, $holders ) {
4588 # Strip bad stuff out of the title (tooltip). We can't just use
4589 # replaceLinkHoldersText() here, because if this function is called
4590 # from replaceInternalLinks2(), mLinkHolders won't be up-to-date.
4592 $tooltip = $holders->replaceText( $caption );
4594 $tooltip = $this->replaceLinkHoldersText( $caption );
4597 # make sure there are no placeholders in thumbnail attributes
4598 # that are later expanded to html- so expand them now and
4600 $tooltip = $this->mStripState
->unstripBoth( $tooltip );
4601 $tooltip = Sanitizer
::stripAllTags( $tooltip );
4607 * Set a flag in the output object indicating that the content is dynamic and
4608 * shouldn't be cached.
4610 function disableCache() {
4611 wfDebug( "Parser output marked as uncacheable.\n" );
4612 $this->mOutput
->mCacheTime
= -1;
4616 * Callback from the Sanitizer for expanding items found in HTML attribute
4617 * values, so they can be safely tested and escaped.
4618 * @param string $text
4619 * @param PPFrame $frame
4623 function attributeStripCallback( &$text, $frame = false ) {
4624 $text = $this->replaceVariables( $text, $frame );
4625 $text = $this->mStripState
->unstripBoth( $text );
4634 function Title( $x = NULL ) { return wfSetVar( $this->mTitle
, $x ); }
4635 function Options( $x = NULL ) { return wfSetVar( $this->mOptions
, $x ); }
4636 function OutputType( $x = NULL ) { return wfSetVar( $this->mOutputType
, $x ); }
4642 function getTags() { return array_merge( array_keys($this->mTransparentTagHooks
), array_keys( $this->mTagHooks
) ); }
4647 * Break wikitext input into sections, and either pull or replace
4648 * some particular section's text.
4650 * External callers should use the getSection and replaceSection methods.
4652 * @param string $text Page wikitext
4653 * @param string $section A section identifier string of the form:
4654 * <flag1> - <flag2> - ... - <section number>
4656 * Currently the only recognised flag is "T", which means the target section number
4657 * was derived during a template inclusion parse, in other words this is a template
4658 * section edit link. If no flags are given, it was an ordinary section edit link.
4659 * This flag is required to avoid a section numbering mismatch when a section is
4660 * enclosed by <includeonly> (bug 6563).
4662 * The section number 0 pulls the text before the first heading; other numbers will
4663 * pull the given section along with its lower-level subsections. If the section is
4664 * not found, $mode=get will return $newtext, and $mode=replace will return $text.
4666 * @param string $mode One of "get" or "replace"
4667 * @param string $newText Replacement text for section data.
4668 * @return string for "get", the extracted section text.
4669 * for "replace", the whole page with the section replaced.
4671 private function extractSections( $text, $section, $mode, $newText='' ) {
4673 $this->clearState();
4674 $this->setTitle( $wgTitle ); // not generally used but removes an ugly failure mode
4675 $this->mOptions
= new ParserOptions
;
4676 $this->setOutputType( self
::OT_WIKI
);
4678 $frame = $this->getPreprocessor()->newFrame();
4680 // Process section extraction flags
4682 $sectionParts = explode( '-', $section );
4683 $sectionIndex = array_pop( $sectionParts );
4684 foreach ( $sectionParts as $part ) {
4685 if ( $part === 'T' ) {
4686 $flags |
= self
::PTD_FOR_INCLUSION
;
4689 // Preprocess the text
4690 $root = $this->preprocessToDom( $text, $flags );
4692 // <h> nodes indicate section breaks
4693 // They can only occur at the top level, so we can find them by iterating the root's children
4694 $node = $root->getFirstChild();
4696 // Find the target section
4697 if ( $sectionIndex == 0 ) {
4698 // Section zero doesn't nest, level=big
4699 $targetLevel = 1000;
4702 if ( $node->getName() === 'h' ) {
4703 $bits = $node->splitHeading();
4704 if ( $bits['i'] == $sectionIndex ) {
4705 $targetLevel = $bits['level'];
4709 if ( $mode === 'replace' ) {
4710 $outText .= $frame->expand( $node, PPFrame
::RECOVER_ORIG
);
4712 $node = $node->getNextSibling();
4718 if ( $mode === 'get' ) {
4725 // Find the end of the section, including nested sections
4727 if ( $node->getName() === 'h' ) {
4728 $bits = $node->splitHeading();
4729 $curLevel = $bits['level'];
4730 if ( $bits['i'] != $sectionIndex && $curLevel <= $targetLevel ) {
4734 if ( $mode === 'get' ) {
4735 $outText .= $frame->expand( $node, PPFrame
::RECOVER_ORIG
);
4737 $node = $node->getNextSibling();
4740 // Write out the remainder (in replace mode only)
4741 if ( $mode === 'replace' ) {
4742 // Output the replacement text
4743 // Add two newlines on -- trailing whitespace in $newText is conventionally
4744 // stripped by the editor, so we need both newlines to restore the paragraph gap
4745 // Only add trailing whitespace if there is newText
4746 if($newText != "") {
4747 $outText .= $newText . "\n\n";
4751 $outText .= $frame->expand( $node, PPFrame
::RECOVER_ORIG
);
4752 $node = $node->getNextSibling();
4756 if ( is_string( $outText ) ) {
4757 // Re-insert stripped tags
4758 $outText = rtrim( $this->mStripState
->unstripBoth( $outText ) );
4765 * This function returns the text of a section, specified by a number ($section).
4766 * A section is text under a heading like == Heading == or \<h1\>Heading\</h1\>, or
4767 * the first section before any such heading (section 0).
4769 * If a section contains subsections, these are also returned.
4771 * @param string $text text to look in
4772 * @param string $section section identifier
4773 * @param string $deftext default to return if section is not found
4774 * @return string text of the requested section
4776 public function getSection( $text, $section, $deftext='' ) {
4777 return $this->extractSections( $text, $section, "get", $deftext );
4780 public function replaceSection( $oldtext, $section, $text ) {
4781 return $this->extractSections( $oldtext, $section, "replace", $text );
4785 * Get the timestamp associated with the current revision, adjusted for
4786 * the default server-local timestamp
4788 function getRevisionTimestamp() {
4789 if ( is_null( $this->mRevisionTimestamp
) ) {
4790 wfProfileIn( __METHOD__
);
4792 $dbr = wfGetDB( DB_SLAVE
);
4793 $timestamp = $dbr->selectField( 'revision', 'rev_timestamp',
4794 array( 'rev_id' => $this->mRevisionId
), __METHOD__
);
4796 // Normalize timestamp to internal MW format for timezone processing.
4797 // This has the added side-effect of replacing a null value with
4798 // the current time, which gives us more sensible behavior for
4800 $timestamp = wfTimestamp( TS_MW
, $timestamp );
4802 // The cryptic '' timezone parameter tells to use the site-default
4803 // timezone offset instead of the user settings.
4805 // Since this value will be saved into the parser cache, served
4806 // to other users, and potentially even used inside links and such,
4807 // it needs to be consistent for all visitors.
4808 $this->mRevisionTimestamp
= $wgContLang->userAdjust( $timestamp, '' );
4810 wfProfileOut( __METHOD__
);
4812 return $this->mRevisionTimestamp
;
4816 * Get the name of the user that edited the last revision
4818 function getRevisionUser() {
4819 // if this template is subst: the revision id will be blank,
4820 // so just use the current user's name
4821 if( $this->mRevisionId
) {
4822 $revision = Revision
::newFromId( $this->mRevisionId
);
4823 $revuser = $revision->getUserText();
4826 $revuser = $wgUser->getName();
4832 * Mutator for $mDefaultSort
4834 * @param $sort New value
4836 public function setDefaultSort( $sort ) {
4837 $this->mDefaultSort
= $sort;
4841 * Accessor for $mDefaultSort
4842 * Will use the title/prefixed title if none is set
4846 public function getDefaultSort() {
4847 global $wgCategoryPrefixedDefaultSortkey;
4848 if( $this->mDefaultSort
!== false ) {
4849 return $this->mDefaultSort
;
4850 } elseif ($this->mTitle
->getNamespace() == NS_CATEGORY ||
4851 !$wgCategoryPrefixedDefaultSortkey) {
4852 return $this->mTitle
->getText();
4854 return $this->mTitle
->getPrefixedText();
4859 * Accessor for $mDefaultSort
4860 * Unlike getDefaultSort(), will return false if none is set
4862 * @return string or false
4864 public function getCustomDefaultSort() {
4865 return $this->mDefaultSort
;
4869 * Try to guess the section anchor name based on a wikitext fragment
4870 * presumably extracted from a heading, for example "Header" from
4873 public function guessSectionNameFromWikiText( $text ) {
4874 # Strip out wikitext links(they break the anchor)
4875 $text = $this->stripSectionName( $text );
4876 $headline = Sanitizer
::decodeCharReferences( $text );
4878 $headline = StringUtils
::delimiterReplace( '<', '>', '', $headline );
4879 $headline = trim( $headline );
4880 $sectionanchor = '#' . urlencode( str_replace( ' ', '_', $headline ) );
4881 $replacearray = array(
4886 array_keys( $replacearray ),
4887 array_values( $replacearray ),
4892 * Strips a text string of wikitext for use in a section anchor
4894 * Accepts a text string and then removes all wikitext from the
4895 * string and leaves only the resultant text (i.e. the result of
4896 * [[User:WikiSysop|Sysop]] would be "Sysop" and the result of
4897 * [[User:WikiSysop]] would be "User:WikiSysop") - this is intended
4898 * to create valid section anchors by mimicing the output of the
4899 * parser when headings are parsed.
4901 * @param $text string Text string to be stripped of wikitext
4902 * for use in a Section anchor
4903 * @return Filtered text string
4905 public function stripSectionName( $text ) {
4906 # Strip internal link markup
4907 $text = preg_replace('/\[\[:?([^[|]+)\|([^[]+)\]\]/','$2',$text);
4908 $text = preg_replace('/\[\[:?([^[]+)\|?\]\]/','$1',$text);
4910 # Strip external link markup (FIXME: Not Tolerant to blank link text
4911 # I.E. [http://www.mediawiki.org] will render as [1] or something depending
4912 # on how many empty links there are on the page - need to figure that out.
4913 $text = preg_replace('/\[(?:' . wfUrlProtocols() . ')([^ ]+?) ([^[]+)\]/','$2',$text);
4915 # Parse wikitext quotes (italics & bold)
4916 $text = $this->doQuotes($text);
4919 $text = StringUtils
::delimiterReplace( '<', '>', '', $text );
4923 function srvus( $text ) {
4924 return $this->testSrvus( $text, $this->mOutputType
);
4928 * strip/replaceVariables/unstrip for preprocessor regression testing
4930 function testSrvus( $text, $title, $options, $outputType = self
::OT_HTML
) {
4931 $this->clearState();
4932 if ( ! ( $title instanceof Title
) ) {
4933 $title = Title
::newFromText( $title );
4935 $this->mTitle
= $title;
4936 $this->mOptions
= $options;
4937 $this->setOutputType( $outputType );
4938 $text = $this->replaceVariables( $text );
4939 $text = $this->mStripState
->unstripBoth( $text );
4940 $text = Sanitizer
::removeHTMLtags( $text );
4944 function testPst( $text, $title, $options ) {
4946 if ( ! ( $title instanceof Title
) ) {
4947 $title = Title
::newFromText( $title );
4949 return $this->preSaveTransform( $text, $title, $wgUser, $options );
4952 function testPreprocess( $text, $title, $options ) {
4953 if ( ! ( $title instanceof Title
) ) {
4954 $title = Title
::newFromText( $title );
4956 return $this->testSrvus( $text, $title, $options, self
::OT_PREPROCESS
);
4959 function markerSkipCallback( $s, $callback ) {
4962 while ( $i < strlen( $s ) ) {
4963 $markerStart = strpos( $s, $this->mUniqPrefix
, $i );
4964 if ( $markerStart === false ) {
4965 $out .= call_user_func( $callback, substr( $s, $i ) );
4968 $out .= call_user_func( $callback, substr( $s, $i, $markerStart - $i ) );
4969 $markerEnd = strpos( $s, self
::MARKER_SUFFIX
, $markerStart );
4970 if ( $markerEnd === false ) {
4971 $out .= substr( $s, $markerStart );
4974 $markerEnd +
= strlen( self
::MARKER_SUFFIX
);
4975 $out .= substr( $s, $markerStart, $markerEnd - $markerStart );
4983 function serialiseHalfParsedText( $text ) {
4985 $data['text'] = $text;
4987 // First, find all strip markers, and store their
4988 // data in an array.
4989 $stripState = new StripState
;
4991 while( ( $start_pos = strpos( $text, $this->mUniqPrefix
, $pos ) ) && ( $end_pos = strpos( $text, self
::MARKER_SUFFIX
, $pos ) ) ) {
4992 $end_pos +
= strlen( self
::MARKER_SUFFIX
);
4993 $marker = substr( $text, $start_pos, $end_pos-$start_pos );
4995 if ( !empty( $this->mStripState
->general
->data
[$marker] ) ) {
4996 $replaceArray = $stripState->general
;
4997 $stripText = $this->mStripState
->general
->data
[$marker];
4998 } elseif ( !empty( $this->mStripState
->nowiki
->data
[$marker] ) ) {
4999 $replaceArray = $stripState->nowiki
;
5000 $stripText = $this->mStripState
->nowiki
->data
[$marker];
5002 throw new MWException( "Hanging strip marker: '$marker'." );
5005 $replaceArray->setPair( $marker, $stripText );
5008 $data['stripstate'] = $stripState;
5010 // Now, find all of our links, and store THEIR
5011 // data in an array! :)
5012 $links = array( 'internal' => array(), 'interwiki' => array() );
5016 while( ( $start_pos = strpos( $text, '<!--LINK ', $pos ) ) ) {
5017 list( $ns, $trail ) = explode( ':', substr( $text, $start_pos +
strlen( '<!--LINK ' ) ), 2 );
5020 if (empty( $links['internal'][$ns] )) {
5021 $links['internal'][$ns] = array();
5024 $key = trim( substr( $trail, 0, strpos( $trail, '-->' ) ) );
5025 $links['internal'][$ns][] = $this->mLinkHolders
->internals
[$ns][$key];
5026 $pos = $start_pos +
strlen( "<!--LINK $ns:$key-->" );
5032 while( ( $start_pos = strpos( $text, '<!--IWLINK ', $pos ) ) ) {
5033 $data = substr( $text, $start_pos );
5034 $key = trim( substr( $data, 0, strpos( $data, '-->' ) ) );
5035 $links['interwiki'][] = $this->mLinkHolders
->interwiki
[$key];
5036 $pos = $start_pos +
strlen( "<!--IWLINK $key-->" );
5039 $data['linkholder'] = $links;
5044 function unserialiseHalfParsedText( $data, $intPrefix = null /* Unique identifying prefix */ ) {
5046 $intPrefix = $this->getRandomString();
5048 // First, extract the strip state.
5049 $stripState = $data['stripstate'];
5050 $this->mStripState
->general
->merge( $stripState->general
);
5051 $this->mStripState
->nowiki
->merge( $stripState->nowiki
);
5053 // Now, extract the text, and renumber links
5054 $text = $data['text'];
5055 $links = $data['linkholder'];
5058 foreach( $links['internal'] as $ns => $nsLinks ) {
5059 foreach( $nsLinks as $key => $entry ) {
5060 $newKey = $intPrefix . '-' . $key;
5061 $this->mLinkHolders
->internals
[$ns][$newKey] = $entry;
5063 $text = str_replace( "<!--LINK $ns:$key-->", "<!--LINK $ns:$newKey-->", $text );
5068 foreach( $links['interwiki'] as $key => $entry ) {
5069 $newKey = "$intPrefix-$key";
5070 $this->mLinkHolders
->interwikis
[$newKey] = $entry;
5072 $text = str_replace( "<!--IWLINK $key-->", "<!--IWLINK $newKey-->", $text );
5075 // Should be good to go.
5081 * @todo document, briefly.
5085 var $general, $nowiki;
5087 function __construct() {
5088 $this->general
= new ReplacementArray
;
5089 $this->nowiki
= new ReplacementArray
;
5092 function unstripGeneral( $text ) {
5093 wfProfileIn( __METHOD__
);
5096 $text = $this->general
->replace( $text );
5097 } while ( $text !== $oldText );
5098 wfProfileOut( __METHOD__
);
5102 function unstripNoWiki( $text ) {
5103 wfProfileIn( __METHOD__
);
5106 $text = $this->nowiki
->replace( $text );
5107 } while ( $text !== $oldText );
5108 wfProfileOut( __METHOD__
);
5112 function unstripBoth( $text ) {
5113 wfProfileIn( __METHOD__
);
5116 $text = $this->general
->replace( $text );
5117 $text = $this->nowiki
->replace( $text );
5118 } while ( $text !== $oldText );
5119 wfProfileOut( __METHOD__
);
5125 * @todo document, briefly.
5128 class OnlyIncludeReplacer
{
5131 function replace( $matches ) {
5132 if ( substr( $matches[1], -1 ) === "\n" ) {
5133 $this->output
.= substr( $matches[1], 0, -1 );
5135 $this->output
.= $matches[1];