5 * 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 four main entry points into the Parser class:
20 * produces HTML output
22 * produces altered wiki markup.
24 * performs brace substitution on MediaWiki messages
26 * removes HTML comments and expands templates
29 * objects: $wgLang, $wgContLang
31 * NOT $wgArticle, $wgUser or $wgTitle. Keep them away!
34 * $wgUseTex*, $wgUseDynamicDates*, $wgInterwikiMagic*,
35 * $wgNamespacesWithSubpages, $wgAllowExternalImages*,
36 * $wgLocaltimezone, $wgAllowSpecialInclusion*,
39 * * only within ParserOptions
47 * Update this version number when the ParserOutput format
48 * changes in an incompatible way, so the parser cache
49 * can automatically discard old data.
51 const VERSION
= '1.6.4';
53 # Flags for Parser::setFunctionHook
54 # Also available as global constants from Defines.php
55 const SFH_NO_HASH
= 1;
56 const SFH_OBJECT_ARGS
= 2;
58 # Constants needed for external link processing
59 # Everything except bracket, space, or control characters
60 const EXT_LINK_URL_CLASS
= '[^][<>"\\x00-\\x20\\x7F]';
61 const EXT_IMAGE_REGEX
= '/^(http:\/\/|https:\/\/)([^][<>"\\x00-\\x20\\x7F]+)
62 \\/([A-Za-z0-9_.,~%\\-+&;#*?!=()@\\x80-\\xFF]+)\\.((?i)gif|png|jpg|jpeg)$/Sx';
64 // State constants for the definition list colon extraction
65 const COLON_STATE_TEXT
= 0;
66 const COLON_STATE_TAG
= 1;
67 const COLON_STATE_TAGSTART
= 2;
68 const COLON_STATE_CLOSETAG
= 3;
69 const COLON_STATE_TAGSLASH
= 4;
70 const COLON_STATE_COMMENT
= 5;
71 const COLON_STATE_COMMENTDASH
= 6;
72 const COLON_STATE_COMMENTDASHDASH
= 7;
74 // Flags for preprocessToDom
75 const PTD_FOR_INCLUSION
= 1;
81 var $mTagHooks, $mTransparentTagHooks, $mFunctionHooks, $mFunctionSynonyms, $mVariables,
82 $mImageParams, $mImageParamsMagicArray, $mStripList, $mMarkerSuffix,
83 $mExtLinkBracketedRegex;
85 # Cleared with clearState():
86 var $mOutput, $mAutonumber, $mDTopen, $mStripState;
87 var $mIncludeCount, $mArgStack, $mLastSection, $mInPre;
88 var $mInterwikiLinkHolders, $mLinkHolders;
89 var $mIncludeSizes, $mPPNodeCount, $mDefaultSort;
90 var $mTplExpandCache,// empty-frame expansion cache
91 $mTemplatePath; // stores an unsorted hash of all the templates already loaded
92 // in this path. Used for loop detection.
93 var $mTplRedirCache, $mTplDomCache, $mHeadings;
96 # These are variables reset at least once per parse regardless of $clearState
97 var $mOptions, // ParserOptions object
98 $mTitle, // Title context, used for self-link rendering and similar things
99 $mOutputType, // Output type, one of the OT_xxx constants
100 $ot, // Shortcut alias, see setOutputType()
101 $mRevisionId, // ID to display in {{REVISIONID}} tags
102 $mRevisionTimestamp, // The timestamp of the specified revision ID
103 $mRevIdForTs; // The revision ID which was used to fetch the timestamp
112 function __construct( $conf = array() ) {
113 $this->mTagHooks
= array();
114 $this->mTransparentTagHooks
= array();
115 $this->mFunctionHooks
= array();
116 $this->mFunctionSynonyms
= array( 0 => array(), 1 => array() );
117 $this->mStripList
= array( 'nowiki', 'gallery' );
118 $this->mMarkerSuffix
= "-QINU\x7f";
119 $this->mExtLinkBracketedRegex
= '/\[(\b(' . wfUrlProtocols() . ')'.
120 '[^][<>"\\x00-\\x20\\x7F]+) *([^\]\\x0a\\x0d]*?)\]/S';
121 $this->mFirstCall
= true;
125 * Do various kinds of initialisation on the first call of the parser
127 function firstCallInit() {
128 if ( !$this->mFirstCall
) {
132 wfProfileIn( __METHOD__
);
133 global $wgAllowDisplayTitle, $wgAllowSlowParserFunctions;
135 $this->setHook( 'pre', array( $this, 'renderPreTag' ) );
137 # Syntax for arguments (see self::setFunctionHook):
138 # "name for lookup in localized magic words array",
140 # optional SFH_NO_HASH to omit the hash from calls (e.g. {{int:...}
141 # instead of {{#int:...}})
142 $this->setFunctionHook( 'int', array( 'CoreParserFunctions', 'intFunction' ), SFH_NO_HASH
);
143 $this->setFunctionHook( 'ns', array( 'CoreParserFunctions', 'ns' ), SFH_NO_HASH
);
144 $this->setFunctionHook( 'urlencode', array( 'CoreParserFunctions', 'urlencode' ), SFH_NO_HASH
);
145 $this->setFunctionHook( 'lcfirst', array( 'CoreParserFunctions', 'lcfirst' ), SFH_NO_HASH
);
146 $this->setFunctionHook( 'ucfirst', array( 'CoreParserFunctions', 'ucfirst' ), SFH_NO_HASH
);
147 $this->setFunctionHook( 'lc', array( 'CoreParserFunctions', 'lc' ), SFH_NO_HASH
);
148 $this->setFunctionHook( 'uc', array( 'CoreParserFunctions', 'uc' ), SFH_NO_HASH
);
149 $this->setFunctionHook( 'localurl', array( 'CoreParserFunctions', 'localurl' ), SFH_NO_HASH
);
150 $this->setFunctionHook( 'localurle', array( 'CoreParserFunctions', 'localurle' ), SFH_NO_HASH
);
151 $this->setFunctionHook( 'fullurl', array( 'CoreParserFunctions', 'fullurl' ), SFH_NO_HASH
);
152 $this->setFunctionHook( 'fullurle', array( 'CoreParserFunctions', 'fullurle' ), SFH_NO_HASH
);
153 $this->setFunctionHook( 'formatnum', array( 'CoreParserFunctions', 'formatnum' ), SFH_NO_HASH
);
154 $this->setFunctionHook( 'grammar', array( 'CoreParserFunctions', 'grammar' ), SFH_NO_HASH
);
155 $this->setFunctionHook( 'plural', array( 'CoreParserFunctions', 'plural' ), SFH_NO_HASH
);
156 $this->setFunctionHook( 'numberofpages', array( 'CoreParserFunctions', 'numberofpages' ), SFH_NO_HASH
);
157 $this->setFunctionHook( 'numberofusers', array( 'CoreParserFunctions', 'numberofusers' ), SFH_NO_HASH
);
158 $this->setFunctionHook( 'numberofarticles', array( 'CoreParserFunctions', 'numberofarticles' ), SFH_NO_HASH
);
159 $this->setFunctionHook( 'numberoffiles', array( 'CoreParserFunctions', 'numberoffiles' ), SFH_NO_HASH
);
160 $this->setFunctionHook( 'numberofadmins', array( 'CoreParserFunctions', 'numberofadmins' ), SFH_NO_HASH
);
161 $this->setFunctionHook( 'numberofedits', array( 'CoreParserFunctions', 'numberofedits' ), SFH_NO_HASH
);
162 $this->setFunctionHook( 'language', array( 'CoreParserFunctions', 'language' ), SFH_NO_HASH
);
163 $this->setFunctionHook( 'padleft', array( 'CoreParserFunctions', 'padleft' ), SFH_NO_HASH
);
164 $this->setFunctionHook( 'padright', array( 'CoreParserFunctions', 'padright' ), SFH_NO_HASH
);
165 $this->setFunctionHook( 'anchorencode', array( 'CoreParserFunctions', 'anchorencode' ), SFH_NO_HASH
);
166 $this->setFunctionHook( 'special', array( 'CoreParserFunctions', 'special' ) );
167 $this->setFunctionHook( 'defaultsort', array( 'CoreParserFunctions', 'defaultsort' ), SFH_NO_HASH
);
168 $this->setFunctionHook( 'filepath', array( 'CoreParserFunctions', 'filepath' ), SFH_NO_HASH
);
170 if ( $wgAllowDisplayTitle ) {
171 $this->setFunctionHook( 'displaytitle', array( 'CoreParserFunctions', 'displaytitle' ), SFH_NO_HASH
);
173 if ( $wgAllowSlowParserFunctions ) {
174 $this->setFunctionHook( 'pagesinnamespace', array( 'CoreParserFunctions', 'pagesinnamespace' ), SFH_NO_HASH
);
177 $this->initialiseVariables();
178 $this->mFirstCall
= false;
179 wfProfileOut( __METHOD__
);
187 function clearState() {
188 wfProfileIn( __METHOD__
);
189 if ( $this->mFirstCall
) {
190 $this->firstCallInit();
192 $this->mOutput
= new ParserOutput
;
193 $this->mAutonumber
= 0;
194 $this->mLastSection
= '';
195 $this->mDTopen
= false;
196 $this->mIncludeCount
= array();
197 $this->mStripState
= new StripState
;
198 $this->mArgStack
= false;
199 $this->mInPre
= false;
200 $this->mInterwikiLinkHolders
= array(
204 $this->mLinkHolders
= array(
205 'namespaces' => array(),
207 'queries' => array(),
211 $this->mRevisionTimestamp
= $this->mRevisionId
= null;
214 * Prefix for temporary replacement strings for the multipass parser.
215 * \x07 should never appear in input as it's disallowed in XML.
216 * Using it at the front also gives us a little extra robustness
217 * since it shouldn't match when butted up against identifier-like
220 * Must not consist of all title characters, or else it will change
221 * the behaviour of <nowiki> in a link.
223 #$this->mUniqPrefix = "\x07UNIQ" . Parser::getRandomString();
224 $this->mUniqPrefix
= "\x7fUNIQ" . Parser
::getRandomString();
226 # Clear these on every parse, bug 4549
227 $this->mTemplatePath
= array();
228 $this->mTplExpandCache
= $this->mTplRedirCache
= $this->mTplDomCache
= array();
230 $this->mShowToc
= true;
231 $this->mForceTocPosition
= false;
232 $this->mIncludeSizes
= array(
236 $this->mPPNodeCount
= 0;
237 $this->mDefaultSort
= false;
238 $this->mHeadings
= array();
240 wfRunHooks( 'ParserClearState', array( &$this ) );
241 wfProfileOut( __METHOD__
);
244 function setOutputType( $ot ) {
245 $this->mOutputType
= $ot;
248 'html' => $ot == OT_HTML
,
249 'wiki' => $ot == OT_WIKI
,
250 'msg' => $ot == OT_MSG
,
251 'pre' => $ot == OT_PREPROCESS
,
256 * Accessor for mUniqPrefix.
260 function uniqPrefix() {
261 if( !isset( $this->mUniqPrefix
) ) {
262 // @fixme this is probably *horribly wrong*
263 // LanguageConverter seems to want $wgParser's uniqPrefix, however
264 // if this is called for a parser cache hit, the parser may not
265 // have ever been initialized in the first place.
266 // Not really sure what the heck is supposed to be going on here.
268 //throw new MWException( "Accessing uninitialized mUniqPrefix" );
270 return $this->mUniqPrefix
;
274 * Convert wikitext to HTML
275 * Do not call this function recursively.
277 * @param string $text Text we want to parse
278 * @param Title &$title A title object
279 * @param array $options
280 * @param boolean $linestart
281 * @param boolean $clearState
282 * @param int $revid number to pass in {{REVISIONID}}
283 * @return ParserOutput a ParserOutput
285 public function parse( $text, &$title, $options, $linestart = true, $clearState = true, $revid = null ) {
287 * First pass--just handle <nowiki> sections, pass the rest off
288 * to internalParse() which does all the real work.
291 global $wgUseTidy, $wgAlwaysUseTidy, $wgContLang;
292 $fname = 'Parser::parse-' . wfGetCaller();
293 wfProfileIn( __METHOD__
);
294 wfProfileIn( $fname );
300 $this->mOptions
= $options;
301 $this->mTitle
=& $title;
302 $oldRevisionId = $this->mRevisionId
;
303 $oldRevisionTimestamp = $this->mRevisionTimestamp
;
304 if( $revid !== null ) {
305 $this->mRevisionId
= $revid;
306 $this->mRevisionTimestamp
= null;
308 $this->setOutputType( OT_HTML
);
309 wfRunHooks( 'ParserBeforeStrip', array( &$this, &$text, &$this->mStripState
) );
311 wfRunHooks( 'ParserAfterStrip', array( &$this, &$text, &$this->mStripState
) );
312 $text = $this->internalParse( $text );
313 $text = $this->mStripState
->unstripGeneral( $text );
315 # Clean up special characters, only run once, next-to-last before doBlockLevels
317 # french spaces, last one Guillemet-left
318 # only if there is something before the space
319 '/(.) (?=\\?|:|;|!|%|\\302\\273)/' => '\\1 \\2',
320 # french spaces, Guillemet-right
321 '/(\\302\\253) /' => '\\1 ',
323 $text = preg_replace( array_keys($fixtags), array_values($fixtags), $text );
326 $text = $this->doBlockLevels( $text, $linestart );
328 $this->replaceLinkHolders( $text );
330 # the position of the parserConvert() call should not be changed. it
331 # assumes that the links are all replaced and the only thing left
332 # is the <nowiki> mark.
333 # Side-effects: this calls $this->mOutput->setTitleText()
334 $text = $wgContLang->parserConvert( $text, $this );
336 $text = $this->mStripState
->unstripNoWiki( $text );
338 wfRunHooks( 'ParserBeforeTidy', array( &$this, &$text ) );
340 //!JF Move to its own function
342 $uniq_prefix = $this->mUniqPrefix
;
344 $elements = array_keys( $this->mTransparentTagHooks
);
345 $text = Parser
::extractTagsAndParams( $elements, $text, $matches, $uniq_prefix );
347 foreach( $matches as $marker => $data ) {
348 list( $element, $content, $params, $tag ) = $data;
349 $tagName = strtolower( $element );
350 if( isset( $this->mTransparentTagHooks
[$tagName] ) ) {
351 $output = call_user_func_array( $this->mTransparentTagHooks
[$tagName],
352 array( $content, $params, $this ) );
356 $this->mStripState
->general
->setPair( $marker, $output );
358 $text = $this->mStripState
->unstripGeneral( $text );
360 $text = Sanitizer
::normalizeCharReferences( $text );
362 if (($wgUseTidy and $this->mOptions
->mTidy
) or $wgAlwaysUseTidy) {
363 $text = Parser
::tidy($text);
365 # attempt to sanitize at least some nesting problems
366 # (bug #2702 and quite a few others)
368 # ''Something [http://www.cool.com cool''] -->
369 # <i>Something</i><a href="http://www.cool.com"..><i>cool></i></a>
370 '/(<([bi])>)(<([bi])>)?([^<]*)(<\/?a[^<]*>)([^<]*)(<\/\\4>)?(<\/\\2>)/' =>
371 '\\1\\3\\5\\8\\9\\6\\1\\3\\7\\8\\9',
372 # fix up an anchor inside another anchor, only
373 # at least for a single single nested link (bug 3695)
374 '/(<a[^>]+>)([^<]*)(<a[^>]+>[^<]*)<\/a>(.*)<\/a>/' =>
375 '\\1\\2</a>\\3</a>\\1\\4</a>',
376 # fix div inside inline elements- doBlockLevels won't wrap a line which
377 # contains a div, so fix it up here; replace
378 # div with escaped text
379 '/(<([aib]) [^>]+>)([^<]*)(<div([^>]*)>)(.*)(<\/div>)([^<]*)(<\/\\2>)/' =>
380 '\\1\\3<div\\5>\\6</div>\\8\\9',
381 # remove empty italic or bold tag pairs, some
382 # introduced by rules above
383 '/<([bi])><\/\\1>/' => '',
386 $text = preg_replace(
387 array_keys( $tidyregs ),
388 array_values( $tidyregs ),
392 wfRunHooks( 'ParserAfterTidy', array( &$this, &$text ) );
394 # Information on include size limits, for the benefit of users who try to skirt them
395 if ( $this->mOptions
->getEnableLimitReport() ) {
396 $max = $this->mOptions
->getMaxIncludeSize();
398 "Preprocessor node count: {$this->mPPNodeCount}/{$this->mOptions->mMaxPPNodeCount}\n" .
399 "Post-expand include size: {$this->mIncludeSizes['post-expand']}/$max bytes\n" .
400 "Template argument size: {$this->mIncludeSizes['arg']}/$max bytes\n";
401 wfRunHooks( 'ParserLimitReport', array( $this, &$limitReport ) );
402 $text .= "\n<!-- \n$limitReport-->\n";
404 $this->mOutput
->setText( $text );
405 $this->mRevisionId
= $oldRevisionId;
406 $this->mRevisionTimestamp
= $oldRevisionTimestamp;
407 wfProfileOut( $fname );
408 wfProfileOut( __METHOD__
);
410 return $this->mOutput
;
414 * Recursive parser entry point that can be called from an extension tag
417 function recursiveTagParse( $text ) {
418 wfProfileIn( __METHOD__
);
419 wfRunHooks( 'ParserBeforeStrip', array( &$this, &$text, &$this->mStripState
) );
420 wfRunHooks( 'ParserAfterStrip', array( &$this, &$text, &$this->mStripState
) );
421 $text = $this->internalParse( $text );
422 wfProfileOut( __METHOD__
);
427 * Expand templates and variables in the text, producing valid, static wikitext.
428 * Also removes comments.
430 function preprocess( $text, $title, $options, $revid = null ) {
431 wfProfileIn( __METHOD__
);
433 $this->setOutputType( OT_PREPROCESS
);
434 $this->mOptions
= $options;
435 $this->mTitle
= $title;
436 if( $revid !== null ) {
437 $this->mRevisionId
= $revid;
439 wfRunHooks( 'ParserBeforeStrip', array( &$this, &$text, &$this->mStripState
) );
440 wfRunHooks( 'ParserAfterStrip', array( &$this, &$text, &$this->mStripState
) );
441 $text = $this->replaceVariables( $text );
442 if ( $this->mOptions
->getRemoveComments() ) {
443 $text = Sanitizer
::removeHTMLcomments( $text );
445 $text = $this->mStripState
->unstripBoth( $text );
446 wfProfileOut( __METHOD__
);
451 * Get a random string
456 function getRandomString() {
457 return dechex(mt_rand(0, 0x7fffffff)) . dechex(mt_rand(0, 0x7fffffff));
460 function &getTitle() { return $this->mTitle
; }
461 function getOptions() { return $this->mOptions
; }
463 function getFunctionLang() {
464 global $wgLang, $wgContLang;
465 return $this->mOptions
->getInterfaceMessage() ?
$wgLang : $wgContLang;
469 * Replaces all occurrences of HTML-style comments and the given tags
470 * in the text with a random marker and returns teh next text. The output
471 * parameter $matches will be an associative array filled with data in
473 * 'UNIQ-xxxxx' => array(
476 * array( 'param' => 'x' ),
477 * '<element param="x">tag content</element>' ) )
479 * @param $elements list of element names. Comments are always extracted.
480 * @param $text Source text string.
481 * @param $uniq_prefix
486 function extractTagsAndParams($elements, $text, &$matches, $uniq_prefix = ''){
491 $taglist = implode( '|', $elements );
492 $start = "/<($taglist)(\\s+[^>]*?|\\s*?)(\/?>)|<(!--)/i";
494 while ( '' != $text ) {
495 $p = preg_split( $start, $text, 2, PREG_SPLIT_DELIM_CAPTURE
);
497 if( count( $p ) < 5 ) {
500 if( count( $p ) > 5 ) {
514 $marker = "$uniq_prefix-$element-" . sprintf('%08X', $n++
) . $this->mMarkerSuffix
;
515 $stripped .= $marker;
517 if ( $close === '/>' ) {
518 // Empty element tag, <tag />
523 if( $element == '!--' ) {
526 $end = "/(<\\/$element\\s*>)/i";
528 $q = preg_split( $end, $inside, 2, PREG_SPLIT_DELIM_CAPTURE
);
530 if( count( $q ) < 3 ) {
531 # No end tag -- let it run out to the end of the text.
540 $matches[$marker] = array( $element,
542 Sanitizer
::decodeTagAttributes( $attributes ),
543 "<$element$attributes$close$content$tail" );
549 * Get a list of strippable XML-like elements
551 function getStripList() {
553 $elements = $this->mStripList
;
555 $elements[] = 'html';
557 if( $this->mOptions
->getUseTeX() ) {
558 $elements[] = 'math';
564 * @deprecated use replaceVariables
566 function strip( $text, $state, $stripcomments = false , $dontstrip = array () ) {
571 * Restores pre, math, and other extensions removed by strip()
573 * always call unstripNoWiki() after this one
575 * @deprecated use $this->mStripState->unstrip()
577 function unstrip( $text, $state ) {
578 return $state->unstripGeneral( $text );
582 * Always call this after unstrip() to preserve the order
585 * @deprecated use $this->mStripState->unstrip()
587 function unstripNoWiki( $text, $state ) {
588 return $state->unstripNoWiki( $text );
592 * @deprecated use $this->mStripState->unstripBoth()
594 function unstripForHTML( $text ) {
595 return $this->mStripState
->unstripBoth( $text );
599 * Add an item to the strip state
600 * Returns the unique tag which must be inserted into the stripped text
601 * The tag will be replaced with the original text in unstrip()
605 function insertStripItem( $text ) {
607 $rnd = "{$this->mUniqPrefix}-item-$n-{$this->mMarkerSuffix}";
609 $this->mStripState
->general
->setPair( $rnd, $text );
614 * Interface with html tidy, used if $wgUseTidy = true.
615 * If tidy isn't able to correct the markup, the original will be
616 * returned in all its glory with a warning comment appended.
618 * Either the external tidy program or the in-process tidy extension
619 * will be used depending on availability. Override the default
620 * $wgTidyInternal setting to disable the internal if it's not working.
622 * @param string $text Hideous HTML input
623 * @return string Corrected HTML output
627 function tidy( $text ) {
628 global $wgTidyInternal;
629 $wrappedtext = '<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN"'.
630 ' "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"><html>'.
631 '<head><title>test</title></head><body>'.$text.'</body></html>';
632 if( $wgTidyInternal ) {
633 $correctedtext = Parser
::internalTidy( $wrappedtext );
635 $correctedtext = Parser
::externalTidy( $wrappedtext );
637 if( is_null( $correctedtext ) ) {
638 wfDebug( "Tidy error detected!\n" );
639 return $text . "\n<!-- Tidy found serious XHTML errors -->\n";
641 return $correctedtext;
645 * Spawn an external HTML tidy process and get corrected markup back from it.
650 function externalTidy( $text ) {
651 global $wgTidyConf, $wgTidyBin, $wgTidyOpts;
652 $fname = 'Parser::externalTidy';
653 wfProfileIn( $fname );
658 $descriptorspec = array(
659 0 => array('pipe', 'r'),
660 1 => array('pipe', 'w'),
661 2 => array('file', wfGetNull(), 'a')
664 $process = proc_open("$wgTidyBin -config $wgTidyConf $wgTidyOpts$opts", $descriptorspec, $pipes);
665 if (is_resource($process)) {
666 // Theoretically, this style of communication could cause a deadlock
667 // here. If the stdout buffer fills up, then writes to stdin could
668 // block. This doesn't appear to happen with tidy, because tidy only
669 // writes to stdout after it's finished reading from stdin. Search
670 // for tidyParseStdin and tidySaveStdout in console/tidy.c
671 fwrite($pipes[0], $text);
673 while (!feof($pipes[1])) {
674 $cleansource .= fgets($pipes[1], 1024);
677 proc_close($process);
680 wfProfileOut( $fname );
682 if( $cleansource == '' && $text != '') {
683 // Some kind of error happened, so we couldn't get the corrected text.
684 // Just give up; we'll use the source text and append a warning.
692 * Use the HTML tidy PECL extension to use the tidy library in-process,
693 * saving the overhead of spawning a new process.
695 * 'pear install tidy' should be able to compile the extension module.
700 function internalTidy( $text ) {
701 global $wgTidyConf, $IP, $wgDebugTidy;
702 $fname = 'Parser::internalTidy';
703 wfProfileIn( $fname );
706 $tidy->parseString( $text, $wgTidyConf, 'utf8' );
707 $tidy->cleanRepair();
708 if( $tidy->getStatus() == 2 ) {
709 // 2 is magic number for fatal error
710 // http://www.php.net/manual/en/function.tidy-get-status.php
713 $cleansource = tidy_get_output( $tidy );
715 if ( $wgDebugTidy && $tidy->getStatus() > 0 ) {
716 $cleansource .= "<!--\nTidy reports:\n" .
717 str_replace( '-->', '-->', $tidy->errorBuffer
) .
721 wfProfileOut( $fname );
726 * parse the wiki syntax used to render tables
730 function doTableStuff ( $text ) {
731 $fname = 'Parser::doTableStuff';
732 wfProfileIn( $fname );
734 $lines = explode ( "\n" , $text );
735 $td_history = array (); // Is currently a td tag open?
736 $last_tag_history = array (); // Save history of last lag activated (td, th or caption)
737 $tr_history = array (); // Is currently a tr tag open?
738 $tr_attributes = array (); // history of tr attributes
739 $has_opened_tr = array(); // Did this table open a <tr> element?
740 $indent_level = 0; // indent level of the table
741 foreach ( $lines as $key => $line )
743 $line = trim ( $line );
745 if( $line == '' ) { // empty line, go to next line
748 $first_character = $line{0};
751 if ( preg_match( '/^(:*)\{\|(.*)$/' , $line , $matches ) ) {
752 // First check if we are starting a new table
753 $indent_level = strlen( $matches[1] );
755 $attributes = $this->mStripState
->unstripBoth( $matches[2] );
756 $attributes = Sanitizer
::fixTagAttributes ( $attributes , 'table' );
758 $lines[$key] = str_repeat( '<dl><dd>' , $indent_level ) . "<table{$attributes}>";
759 array_push ( $td_history , false );
760 array_push ( $last_tag_history , '' );
761 array_push ( $tr_history , false );
762 array_push ( $tr_attributes , '' );
763 array_push ( $has_opened_tr , false );
764 } else if ( count ( $td_history ) == 0 ) {
765 // Don't do any of the following
767 } else if ( substr ( $line , 0 , 2 ) == '|}' ) {
768 // We are ending a table
769 $line = '</table>' . substr ( $line , 2 );
770 $last_tag = array_pop ( $last_tag_history );
772 if ( !array_pop ( $has_opened_tr ) ) {
773 $line = "<tr><td></td></tr>{$line}";
776 if ( array_pop ( $tr_history ) ) {
777 $line = "</tr>{$line}";
780 if ( array_pop ( $td_history ) ) {
781 $line = "</{$last_tag}>{$line}";
783 array_pop ( $tr_attributes );
784 $lines[$key] = $line . str_repeat( '</dd></dl>' , $indent_level );
785 } else if ( substr ( $line , 0 , 2 ) == '|-' ) {
786 // Now we have a table row
787 $line = preg_replace( '#^\|-+#', '', $line );
789 // Whats after the tag is now only attributes
790 $attributes = $this->mStripState
->unstripBoth( $line );
791 $attributes = Sanitizer
::fixTagAttributes ( $attributes , 'tr' );
792 array_pop ( $tr_attributes );
793 array_push ( $tr_attributes , $attributes );
796 $last_tag = array_pop ( $last_tag_history );
797 array_pop ( $has_opened_tr );
798 array_push ( $has_opened_tr , true );
800 if ( array_pop ( $tr_history ) ) {
804 if ( array_pop ( $td_history ) ) {
805 $line = "</{$last_tag}>{$line}";
808 $lines[$key] = $line;
809 array_push ( $tr_history , false );
810 array_push ( $td_history , false );
811 array_push ( $last_tag_history , '' );
813 else if ( $first_character == '|' ||
$first_character == '!' ||
substr ( $line , 0 , 2 ) == '|+' ) {
814 // This might be cell elements, td, th or captions
815 if ( substr ( $line , 0 , 2 ) == '|+' ) {
816 $first_character = '+';
817 $line = substr ( $line , 1 );
820 $line = substr ( $line , 1 );
822 if ( $first_character == '!' ) {
823 $line = str_replace ( '!!' , '||' , $line );
826 // Split up multiple cells on the same line.
827 // FIXME : This can result in improper nesting of tags processed
828 // by earlier parser steps, but should avoid splitting up eg
829 // attribute values containing literal "||".
830 $cells = StringUtils
::explodeMarkup( '||' , $line );
834 // Loop through each table cell
835 foreach ( $cells as $cell )
838 if ( $first_character != '+' )
840 $tr_after = array_pop ( $tr_attributes );
841 if ( !array_pop ( $tr_history ) ) {
842 $previous = "<tr{$tr_after}>\n";
844 array_push ( $tr_history , true );
845 array_push ( $tr_attributes , '' );
846 array_pop ( $has_opened_tr );
847 array_push ( $has_opened_tr , true );
850 $last_tag = array_pop ( $last_tag_history );
852 if ( array_pop ( $td_history ) ) {
853 $previous = "</{$last_tag}>{$previous}";
856 if ( $first_character == '|' ) {
858 } else if ( $first_character == '!' ) {
860 } else if ( $first_character == '+' ) {
861 $last_tag = 'caption';
866 array_push ( $last_tag_history , $last_tag );
868 // A cell could contain both parameters and data
869 $cell_data = explode ( '|' , $cell , 2 );
871 // Bug 553: Note that a '|' inside an invalid link should not
872 // be mistaken as delimiting cell parameters
873 if ( strpos( $cell_data[0], '[[' ) !== false ) {
874 $cell = "{$previous}<{$last_tag}>{$cell}";
875 } else if ( count ( $cell_data ) == 1 )
876 $cell = "{$previous}<{$last_tag}>{$cell_data[0]}";
878 $attributes = $this->mStripState
->unstripBoth( $cell_data[0] );
879 $attributes = Sanitizer
::fixTagAttributes( $attributes , $last_tag );
880 $cell = "{$previous}<{$last_tag}{$attributes}>{$cell_data[1]}";
883 $lines[$key] .= $cell;
884 array_push ( $td_history , true );
889 // Closing open td, tr && table
890 while ( count ( $td_history ) > 0 )
892 if ( array_pop ( $td_history ) ) {
895 if ( array_pop ( $tr_history ) ) {
898 if ( !array_pop ( $has_opened_tr ) ) {
899 $lines[] = "<tr><td></td></tr>" ;
902 $lines[] = '</table>' ;
905 $output = implode ( "\n" , $lines ) ;
907 // special case: don't return empty table
908 if( $output == "<table>\n<tr><td></td></tr>\n</table>" ) {
912 wfProfileOut( $fname );
918 * Helper function for parse() that transforms wiki markup into
919 * HTML. Only called for $mOutputType == OT_HTML.
923 function internalParse( $text ) {
925 $fname = 'Parser::internalParse';
926 wfProfileIn( $fname );
928 # Hook to suspend the parser in this state
929 if ( !wfRunHooks( 'ParserBeforeInternalParse', array( &$this, &$text, &$this->mStripState
) ) ) {
930 wfProfileOut( $fname );
934 $text = $this->replaceVariables( $text );
935 $text = Sanitizer
::removeHTMLtags( $text, array( &$this, 'attributeStripCallback' ), false, array_keys( $this->mTransparentTagHooks
) );
936 wfRunHooks( 'InternalParseBeforeLinks', array( &$this, &$text, &$this->mStripState
) );
938 // Tables need to come after variable replacement for things to work
939 // properly; putting them before other transformations should keep
940 // exciting things like link expansions from showing up in surprising
942 $text = $this->doTableStuff( $text );
944 $text = preg_replace( '/(^|\n)-----*/', '\\1<hr />', $text );
946 $text = $this->stripToc( $text );
947 $this->stripNoGallery( $text );
948 $text = $this->doHeadings( $text );
949 if($this->mOptions
->getUseDynamicDates()) {
950 $df =& DateFormatter
::getInstance();
951 $text = $df->reformat( $this->mOptions
->getDateFormat(), $text );
953 $text = $this->doAllQuotes( $text );
954 $text = $this->replaceInternalLinks( $text );
955 $text = $this->replaceExternalLinks( $text );
957 # replaceInternalLinks may sometimes leave behind
958 # absolute URLs, which have to be masked to hide them from replaceExternalLinks
959 $text = str_replace($this->mUniqPrefix
."NOPARSE", "", $text);
961 $text = $this->doMagicLinks( $text );
962 $text = $this->formatHeadings( $text, $isMain );
964 wfProfileOut( $fname );
969 * Replace special strings like "ISBN xxx" and "RFC xxx" with
970 * magic external links.
974 function doMagicLinks( $text ) {
975 wfProfileIn( __METHOD__
);
976 $text = preg_replace_callback(
978 <a.*?</a> | # Skip link text
979 <.*?> | # Skip stuff inside HTML elements
980 (?:RFC|PMID)\s+([0-9]+) | # RFC or PMID, capture number as m[1]
981 ISBN\s+(\b # ISBN, capture number as m[2]
982 (?: 97[89] [\ \-]? )? # optional 13-digit ISBN prefix
983 (?: [0-9] [\ \-]? ){9} # 9 digits with opt. delimiters
984 [0-9Xx] # check digit
986 )!x', array( &$this, 'magicLinkCallback' ), $text );
987 wfProfileOut( __METHOD__
);
991 function magicLinkCallback( $m ) {
992 if ( substr( $m[0], 0, 1 ) == '<' ) {
995 } elseif ( substr( $m[0], 0, 4 ) == 'ISBN' ) {
997 $num = strtr( $isbn, array(
1002 $titleObj = SpecialPage
::getTitleFor( 'Booksources' );
1003 $text = '<a href="' .
1004 $titleObj->escapeLocalUrl( "isbn=$num" ) .
1005 "\" class=\"internal\">ISBN $isbn</a>";
1007 if ( substr( $m[0], 0, 3 ) == 'RFC' ) {
1011 } elseif ( substr( $m[0], 0, 4 ) == 'PMID' ) {
1013 $urlmsg = 'pubmedurl';
1016 throw new MWException( __METHOD__
.': unrecognised match type "' .
1017 substr($m[0], 0, 20 ) . '"' );
1020 $url = wfMsg( $urlmsg, $id);
1021 $sk = $this->mOptions
->getSkin();
1022 $la = $sk->getExternalLinkAttributes( $url, $keyword.$id );
1023 $text = "<a href=\"{$url}\"{$la}>{$keyword} {$id}</a>";
1029 * Parse headers and return html
1033 function doHeadings( $text ) {
1034 $fname = 'Parser::doHeadings';
1035 wfProfileIn( $fname );
1036 for ( $i = 6; $i >= 1; --$i ) {
1037 $h = str_repeat( '=', $i );
1038 $text = preg_replace( "/^$h(.+)$h\\s*$/m",
1039 "<h$i>\\1</h$i>", $text );
1041 wfProfileOut( $fname );
1046 * Replace single quotes with HTML markup
1048 * @return string the altered text
1050 function doAllQuotes( $text ) {
1051 $fname = 'Parser::doAllQuotes';
1052 wfProfileIn( $fname );
1054 $lines = explode( "\n", $text );
1055 foreach ( $lines as $line ) {
1056 $outtext .= $this->doQuotes ( $line ) . "\n";
1058 $outtext = substr($outtext, 0,-1);
1059 wfProfileOut( $fname );
1064 * Helper function for doAllQuotes()
1066 public function doQuotes( $text ) {
1067 $arr = preg_split( "/(''+)/", $text, -1, PREG_SPLIT_DELIM_CAPTURE
);
1068 if ( count( $arr ) == 1 )
1072 # First, do some preliminary work. This may shift some apostrophes from
1073 # being mark-up to being text. It also counts the number of occurrences
1074 # of bold and italics mark-ups.
1078 foreach ( $arr as $r )
1080 if ( ( $i %
2 ) == 1 )
1082 # If there are ever four apostrophes, assume the first is supposed to
1083 # be text, and the remaining three constitute mark-up for bold text.
1084 if ( strlen( $arr[$i] ) == 4 )
1089 # If there are more than 5 apostrophes in a row, assume they're all
1090 # text except for the last 5.
1091 else if ( strlen( $arr[$i] ) > 5 )
1093 $arr[$i-1] .= str_repeat( "'", strlen( $arr[$i] ) - 5 );
1096 # Count the number of occurrences of bold and italics mark-ups.
1097 # We are not counting sequences of five apostrophes.
1098 if ( strlen( $arr[$i] ) == 2 ) { $numitalics++
; }
1099 else if ( strlen( $arr[$i] ) == 3 ) { $numbold++
; }
1100 else if ( strlen( $arr[$i] ) == 5 ) { $numitalics++
; $numbold++
; }
1105 # If there is an odd number of both bold and italics, it is likely
1106 # that one of the bold ones was meant to be an apostrophe followed
1107 # by italics. Which one we cannot know for certain, but it is more
1108 # likely to be one that has a single-letter word before it.
1109 if ( ( $numbold %
2 == 1 ) && ( $numitalics %
2 == 1 ) )
1112 $firstsingleletterword = -1;
1113 $firstmultiletterword = -1;
1115 foreach ( $arr as $r )
1117 if ( ( $i %
2 == 1 ) and ( strlen( $r ) == 3 ) )
1119 $x1 = substr ($arr[$i-1], -1);
1120 $x2 = substr ($arr[$i-1], -2, 1);
1122 if ($firstspace == -1) $firstspace = $i;
1123 } else if ($x2 == ' ') {
1124 if ($firstsingleletterword == -1) $firstsingleletterword = $i;
1126 if ($firstmultiletterword == -1) $firstmultiletterword = $i;
1132 # If there is a single-letter word, use it!
1133 if ($firstsingleletterword > -1)
1135 $arr [ $firstsingleletterword ] = "''";
1136 $arr [ $firstsingleletterword-1 ] .= "'";
1138 # If not, but there's a multi-letter word, use that one.
1139 else if ($firstmultiletterword > -1)
1141 $arr [ $firstmultiletterword ] = "''";
1142 $arr [ $firstmultiletterword-1 ] .= "'";
1144 # ... otherwise use the first one that has neither.
1145 # (notice that it is possible for all three to be -1 if, for example,
1146 # there is only one pentuple-apostrophe in the line)
1147 else if ($firstspace > -1)
1149 $arr [ $firstspace ] = "''";
1150 $arr [ $firstspace-1 ] .= "'";
1154 # Now let's actually convert our apostrophic mush to HTML!
1159 foreach ($arr as $r)
1163 if ($state == 'both')
1170 if (strlen ($r) == 2)
1173 { $output .= '</i>'; $state = ''; }
1174 else if ($state == 'bi')
1175 { $output .= '</i>'; $state = 'b'; }
1176 else if ($state == 'ib')
1177 { $output .= '</b></i><b>'; $state = 'b'; }
1178 else if ($state == 'both')
1179 { $output .= '<b><i>'.$buffer.'</i>'; $state = 'b'; }
1180 else # $state can be 'b' or ''
1181 { $output .= '<i>'; $state .= 'i'; }
1183 else if (strlen ($r) == 3)
1186 { $output .= '</b>'; $state = ''; }
1187 else if ($state == 'bi')
1188 { $output .= '</i></b><i>'; $state = 'i'; }
1189 else if ($state == 'ib')
1190 { $output .= '</b>'; $state = 'i'; }
1191 else if ($state == 'both')
1192 { $output .= '<i><b>'.$buffer.'</b>'; $state = 'i'; }
1193 else # $state can be 'i' or ''
1194 { $output .= '<b>'; $state .= 'b'; }
1196 else if (strlen ($r) == 5)
1199 { $output .= '</b><i>'; $state = 'i'; }
1200 else if ($state == 'i')
1201 { $output .= '</i><b>'; $state = 'b'; }
1202 else if ($state == 'bi')
1203 { $output .= '</i></b>'; $state = ''; }
1204 else if ($state == 'ib')
1205 { $output .= '</b></i>'; $state = ''; }
1206 else if ($state == 'both')
1207 { $output .= '<i><b>'.$buffer.'</b></i>'; $state = ''; }
1208 else # ($state == '')
1209 { $buffer = ''; $state = 'both'; }
1214 # Now close all remaining tags. Notice that the order is important.
1215 if ($state == 'b' ||
$state == 'ib')
1217 if ($state == 'i' ||
$state == 'bi' ||
$state == 'ib')
1221 # There might be lonely ''''', so make sure we have a buffer
1222 if ($state == 'both' && $buffer)
1223 $output .= '<b><i>'.$buffer.'</i></b>';
1229 * Replace external links
1231 * Note: this is all very hackish and the order of execution matters a lot.
1232 * Make sure to run maintenance/parserTests.php if you change this code.
1236 function replaceExternalLinks( $text ) {
1238 $fname = 'Parser::replaceExternalLinks';
1239 wfProfileIn( $fname );
1241 $sk = $this->mOptions
->getSkin();
1243 $bits = preg_split( $this->mExtLinkBracketedRegex
, $text, -1, PREG_SPLIT_DELIM_CAPTURE
);
1245 $s = $this->replaceFreeExternalLinks( array_shift( $bits ) );
1248 while ( $i<count( $bits ) ) {
1250 $protocol = $bits[$i++
];
1251 $text = $bits[$i++
];
1252 $trail = $bits[$i++
];
1254 # The characters '<' and '>' (which were escaped by
1255 # removeHTMLtags()) should not be included in
1256 # URLs, per RFC 2396.
1258 if (preg_match('/&(lt|gt);/', $url, $m2, PREG_OFFSET_CAPTURE
)) {
1259 $text = substr($url, $m2[0][1]) . ' ' . $text;
1260 $url = substr($url, 0, $m2[0][1]);
1263 # If the link text is an image URL, replace it with an <img> tag
1264 # This happened by accident in the original parser, but some people used it extensively
1265 $img = $this->maybeMakeExternalImage( $text );
1266 if ( $img !== false ) {
1272 # Set linktype for CSS - if URL==text, link is essentially free
1273 $linktype = ($text == $url) ?
'free' : 'text';
1275 # No link text, e.g. [http://domain.tld/some.link]
1276 if ( $text == '' ) {
1277 # Autonumber if allowed. See bug #5918
1278 if ( strpos( wfUrlProtocols(), substr($protocol, 0, strpos($protocol, ':')) ) !== false ) {
1279 $text = '[' . ++
$this->mAutonumber
. ']';
1280 $linktype = 'autonumber';
1282 # Otherwise just use the URL
1283 $text = htmlspecialchars( $url );
1287 # Have link text, e.g. [http://domain.tld/some.link text]s
1289 list( $dtrail, $trail ) = Linker
::splitTrail( $trail );
1292 $text = $wgContLang->markNoConversion($text);
1294 $url = Sanitizer
::cleanUrl( $url );
1296 # Process the trail (i.e. everything after this link up until start of the next link),
1297 # replacing any non-bracketed links
1298 $trail = $this->replaceFreeExternalLinks( $trail );
1300 # Use the encoded URL
1301 # This means that users can paste URLs directly into the text
1302 # Funny characters like ö aren't valid in URLs anyway
1303 # This was changed in August 2004
1304 $s .= $sk->makeExternalLink( $url, $text, false, $linktype, $this->mTitle
->getNamespace() ) . $dtrail . $trail;
1306 # Register link in the output object.
1307 # Replace unnecessary URL escape codes with the referenced character
1308 # This prevents spammers from hiding links from the filters
1309 $pasteurized = Parser
::replaceUnusualEscapes( $url );
1310 $this->mOutput
->addExternalLink( $pasteurized );
1313 wfProfileOut( $fname );
1318 * Replace anything that looks like a URL with a link
1321 function replaceFreeExternalLinks( $text ) {
1323 $fname = 'Parser::replaceFreeExternalLinks';
1324 wfProfileIn( $fname );
1326 $bits = preg_split( '/(\b(?:' . wfUrlProtocols() . '))/S', $text, -1, PREG_SPLIT_DELIM_CAPTURE
);
1327 $s = array_shift( $bits );
1330 $sk = $this->mOptions
->getSkin();
1332 while ( $i < count( $bits ) ){
1333 $protocol = $bits[$i++
];
1334 $remainder = $bits[$i++
];
1337 if ( preg_match( '/^('.self
::EXT_LINK_URL_CLASS
.'+)(.*)$/s', $remainder, $m ) ) {
1338 # Found some characters after the protocol that look promising
1339 $url = $protocol . $m[1];
1342 # special case: handle urls as url args:
1343 # http://www.example.com/foo?=http://www.example.com/bar
1344 if(strlen($trail) == 0 &&
1346 preg_match('/^'. wfUrlProtocols() . '$/S', $bits[$i]) &&
1347 preg_match( '/^('.self
::EXT_LINK_URL_CLASS
.'+)(.*)$/s', $bits[$i +
1], $m ))
1350 $url .= $bits[$i] . $m[1]; # protocol, url as arg to previous link
1355 # The characters '<' and '>' (which were escaped by
1356 # removeHTMLtags()) should not be included in
1357 # URLs, per RFC 2396.
1359 if (preg_match('/&(lt|gt);/', $url, $m2, PREG_OFFSET_CAPTURE
)) {
1360 $trail = substr($url, $m2[0][1]) . $trail;
1361 $url = substr($url, 0, $m2[0][1]);
1364 # Move trailing punctuation to $trail
1366 # If there is no left bracket, then consider right brackets fair game too
1367 if ( strpos( $url, '(' ) === false ) {
1371 $numSepChars = strspn( strrev( $url ), $sep );
1372 if ( $numSepChars ) {
1373 $trail = substr( $url, -$numSepChars ) . $trail;
1374 $url = substr( $url, 0, -$numSepChars );
1377 $url = Sanitizer
::cleanUrl( $url );
1379 # Is this an external image?
1380 $text = $this->maybeMakeExternalImage( $url );
1381 if ( $text === false ) {
1382 # Not an image, make a link
1383 $text = $sk->makeExternalLink( $url, $wgContLang->markNoConversion($url), true, 'free', $this->mTitle
->getNamespace() );
1384 # Register it in the output object...
1385 # Replace unnecessary URL escape codes with their equivalent characters
1386 $pasteurized = Parser
::replaceUnusualEscapes( $url );
1387 $this->mOutput
->addExternalLink( $pasteurized );
1389 $s .= $text . $trail;
1391 $s .= $protocol . $remainder;
1394 wfProfileOut( $fname );
1399 * Replace unusual URL escape codes with their equivalent characters
1403 * @todo This can merge genuinely required bits in the path or query string,
1404 * breaking legit URLs. A proper fix would treat the various parts of
1405 * the URL differently; as a workaround, just use the output for
1406 * statistical records, not for actual linking/output.
1408 static function replaceUnusualEscapes( $url ) {
1409 return preg_replace_callback( '/%[0-9A-Fa-f]{2}/',
1410 array( 'Parser', 'replaceUnusualEscapesCallback' ), $url );
1414 * Callback function used in replaceUnusualEscapes().
1415 * Replaces unusual URL escape codes with their equivalent character
1419 private static function replaceUnusualEscapesCallback( $matches ) {
1420 $char = urldecode( $matches[0] );
1421 $ord = ord( $char );
1422 // Is it an unsafe or HTTP reserved character according to RFC 1738?
1423 if ( $ord > 32 && $ord < 127 && strpos( '<>"#{}|\^~[]`;/?', $char ) === false ) {
1424 // No, shouldn't be escaped
1427 // Yes, leave it escaped
1433 * make an image if it's allowed, either through the global
1434 * option or through the exception
1437 function maybeMakeExternalImage( $url ) {
1438 $sk = $this->mOptions
->getSkin();
1439 $imagesfrom = $this->mOptions
->getAllowExternalImagesFrom();
1440 $imagesexception = !empty($imagesfrom);
1442 if ( $this->mOptions
->getAllowExternalImages()
1443 ||
( $imagesexception && strpos( $url, $imagesfrom ) === 0 ) ) {
1444 if ( preg_match( self
::EXT_IMAGE_REGEX
, $url ) ) {
1446 $text = $sk->makeExternalImage( htmlspecialchars( $url ) );
1453 * Process [[ ]] wikilinks
1457 function replaceInternalLinks( $s ) {
1459 static $fname = 'Parser::replaceInternalLinks' ;
1461 wfProfileIn( $fname );
1463 wfProfileIn( $fname.'-setup' );
1465 # the % is needed to support urlencoded titles as well
1466 if ( !$tc ) { $tc = Title
::legalChars() . '#%'; }
1468 $sk = $this->mOptions
->getSkin();
1470 #split the entire text string on occurences of [[
1471 $a = explode( '[[', ' ' . $s );
1472 #get the first element (all text up to first [[), and remove the space we added
1473 $s = array_shift( $a );
1474 $s = substr( $s, 1 );
1476 # Match a link having the form [[namespace:link|alternate]]trail
1478 if ( !$e1 ) { $e1 = "/^([{$tc}]+)(?:\\|(.+?))?]](.*)\$/sD"; }
1479 # Match cases where there is no "]]", which might still be images
1480 static $e1_img = FALSE;
1481 if ( !$e1_img ) { $e1_img = "/^([{$tc}]+)\\|(.*)\$/sD"; }
1483 $useLinkPrefixExtension = $wgContLang->linkPrefixExtension();
1485 if ( $useLinkPrefixExtension ) {
1486 # Match the end of a line for a word that's not followed by whitespace,
1487 # e.g. in the case of 'The Arab al[[Razi]]', 'al' will be matched
1488 $e2 = wfMsgForContent( 'linkprefix' );
1491 if( is_null( $this->mTitle
) ) {
1492 throw new MWException( __METHOD__
.": \$this->mTitle is null\n" );
1494 $nottalk = !$this->mTitle
->isTalkPage();
1496 if ( $useLinkPrefixExtension ) {
1498 if ( preg_match( $e2, $s, $m ) ) {
1499 $first_prefix = $m[2];
1501 $first_prefix = false;
1507 if($wgContLang->hasVariants()) {
1508 $selflink = $wgContLang->convertLinkToAllVariants($this->mTitle
->getPrefixedText());
1510 $selflink = array($this->mTitle
->getPrefixedText());
1512 $useSubpages = $this->areSubpagesAllowed();
1513 wfProfileOut( $fname.'-setup' );
1515 # Loop for each link
1516 for ($k = 0; isset( $a[$k] ); $k++
) {
1518 if ( $useLinkPrefixExtension ) {
1519 wfProfileIn( $fname.'-prefixhandling' );
1520 if ( preg_match( $e2, $s, $m ) ) {
1528 $prefix = $first_prefix;
1529 $first_prefix = false;
1531 wfProfileOut( $fname.'-prefixhandling' );
1534 $might_be_img = false;
1536 wfProfileIn( "$fname-e1" );
1537 if ( preg_match( $e1, $line, $m ) ) { # page with normal text or alt
1539 # If we get a ] at the beginning of $m[3] that means we have a link that's something like:
1540 # [[Image:Foo.jpg|[http://example.com desc]]] <- having three ] in a row fucks up,
1541 # the real problem is with the $e1 regex
1544 # Still some problems for cases where the ] is meant to be outside punctuation,
1545 # and no image is in sight. See bug 2095.
1548 substr( $m[3], 0, 1 ) === ']' &&
1549 strpos($text, '[') !== false
1552 $text .= ']'; # so that replaceExternalLinks($text) works later
1553 $m[3] = substr( $m[3], 1 );
1555 # fix up urlencoded title texts
1556 if( strpos( $m[1], '%' ) !== false ) {
1557 # Should anchors '#' also be rejected?
1558 $m[1] = str_replace( array('<', '>'), array('<', '>'), urldecode($m[1]) );
1561 } elseif( preg_match($e1_img, $line, $m) ) { # Invalid, but might be an image with a link in its caption
1562 $might_be_img = true;
1564 if ( strpos( $m[1], '%' ) !== false ) {
1565 $m[1] = urldecode($m[1]);
1568 } else { # Invalid form; output directly
1569 $s .= $prefix . '[[' . $line ;
1570 wfProfileOut( "$fname-e1" );
1573 wfProfileOut( "$fname-e1" );
1574 wfProfileIn( "$fname-misc" );
1576 # Don't allow internal links to pages containing
1577 # PROTO: where PROTO is a valid URL protocol; these
1578 # should be external links.
1579 if (preg_match('/^\b(?:' . wfUrlProtocols() . ')/', $m[1])) {
1580 $s .= $prefix . '[[' . $line ;
1584 # Make subpage if necessary
1585 if( $useSubpages ) {
1586 $link = $this->maybeDoSubpageLink( $m[1], $text );
1591 $noforce = (substr($m[1], 0, 1) != ':');
1593 # Strip off leading ':'
1594 $link = substr($link, 1);
1597 wfProfileOut( "$fname-misc" );
1598 wfProfileIn( "$fname-title" );
1599 $nt = Title
::newFromText( $this->mStripState
->unstripNoWiki($link) );
1601 $s .= $prefix . '[[' . $line;
1602 wfProfileOut( "$fname-title" );
1606 $ns = $nt->getNamespace();
1607 $iw = $nt->getInterWiki();
1608 wfProfileOut( "$fname-title" );
1610 if ($might_be_img) { # if this is actually an invalid link
1611 wfProfileIn( "$fname-might_be_img" );
1612 if ($ns == NS_IMAGE
&& $noforce) { #but might be an image
1614 while (isset ($a[$k+
1]) ) {
1615 #look at the next 'line' to see if we can close it there
1616 $spliced = array_splice( $a, $k +
1, 1 );
1617 $next_line = array_shift( $spliced );
1618 $m = explode( ']]', $next_line, 3 );
1619 if ( count( $m ) == 3 ) {
1620 # the first ]] closes the inner link, the second the image
1622 $text .= "[[{$m[0]}]]{$m[1]}";
1625 } elseif ( count( $m ) == 2 ) {
1626 #if there's exactly one ]] that's fine, we'll keep looking
1627 $text .= "[[{$m[0]}]]{$m[1]}";
1629 #if $next_line is invalid too, we need look no further
1630 $text .= '[[' . $next_line;
1635 # we couldn't find the end of this imageLink, so output it raw
1636 #but don't ignore what might be perfectly normal links in the text we've examined
1637 $text = $this->replaceInternalLinks($text);
1638 $s .= "{$prefix}[[$link|$text";
1639 # note: no $trail, because without an end, there *is* no trail
1640 wfProfileOut( "$fname-might_be_img" );
1643 } else { #it's not an image, so output it raw
1644 $s .= "{$prefix}[[$link|$text";
1645 # note: no $trail, because without an end, there *is* no trail
1646 wfProfileOut( "$fname-might_be_img" );
1649 wfProfileOut( "$fname-might_be_img" );
1652 $wasblank = ( '' == $text );
1653 if( $wasblank ) $text = $link;
1655 # Link not escaped by : , create the various objects
1659 wfProfileIn( "$fname-interwiki" );
1660 if( $iw && $this->mOptions
->getInterwikiMagic() && $nottalk && $wgContLang->getLanguageName( $iw ) ) {
1661 $this->mOutput
->addLanguageLink( $nt->getFullText() );
1662 $s = rtrim($s . $prefix);
1663 $s .= trim($trail, "\n") == '' ?
'': $prefix . $trail;
1664 wfProfileOut( "$fname-interwiki" );
1667 wfProfileOut( "$fname-interwiki" );
1669 if ( $ns == NS_IMAGE
) {
1670 wfProfileIn( "$fname-image" );
1671 if ( !wfIsBadImage( $nt->getDBkey(), $this->mTitle
) ) {
1672 # recursively parse links inside the image caption
1673 # actually, this will parse them in any other parameters, too,
1674 # but it might be hard to fix that, and it doesn't matter ATM
1675 $text = $this->replaceExternalLinks($text);
1676 $text = $this->replaceInternalLinks($text);
1678 # cloak any absolute URLs inside the image markup, so replaceExternalLinks() won't touch them
1679 $s .= $prefix . $this->armorLinks( $this->makeImage( $nt, $text ) ) . $trail;
1680 $this->mOutput
->addImage( $nt->getDBkey() );
1682 wfProfileOut( "$fname-image" );
1685 # We still need to record the image's presence on the page
1686 $this->mOutput
->addImage( $nt->getDBkey() );
1688 wfProfileOut( "$fname-image" );
1692 if ( $ns == NS_CATEGORY
) {
1693 wfProfileIn( "$fname-category" );
1694 $s = rtrim($s . "\n"); # bug 87
1697 $sortkey = $this->getDefaultSort();
1701 $sortkey = Sanitizer
::decodeCharReferences( $sortkey );
1702 $sortkey = str_replace( "\n", '', $sortkey );
1703 $sortkey = $wgContLang->convertCategoryKey( $sortkey );
1704 $this->mOutput
->addCategory( $nt->getDBkey(), $sortkey );
1707 * Strip the whitespace Category links produce, see bug 87
1708 * @todo We might want to use trim($tmp, "\n") here.
1710 $s .= trim($prefix . $trail, "\n") == '' ?
'': $prefix . $trail;
1712 wfProfileOut( "$fname-category" );
1717 # Self-link checking
1718 if( $nt->getFragment() === '' ) {
1719 if( in_array( $nt->getPrefixedText(), $selflink, true ) ) {
1720 $s .= $prefix . $sk->makeSelfLinkObj( $nt, $text, '', $trail );
1725 # Special and Media are pseudo-namespaces; no pages actually exist in them
1726 if( $ns == NS_MEDIA
) {
1727 $link = $sk->makeMediaLinkObj( $nt, $text );
1728 # Cloak with NOPARSE to avoid replacement in replaceExternalLinks
1729 $s .= $prefix . $this->armorLinks( $link ) . $trail;
1730 $this->mOutput
->addImage( $nt->getDBkey() );
1732 } elseif( $ns == NS_SPECIAL
) {
1733 if( SpecialPage
::exists( $nt->getDBkey() ) ) {
1734 $s .= $this->makeKnownLinkHolder( $nt, $text, '', $trail, $prefix );
1736 $s .= $this->makeLinkHolder( $nt, $text, '', $trail, $prefix );
1739 } elseif( $ns == NS_IMAGE
) {
1740 $img = wfFindFile( $nt );
1742 // Force a blue link if the file exists; may be a remote
1743 // upload on the shared repository, and we want to see its
1744 // auto-generated page.
1745 $s .= $this->makeKnownLinkHolder( $nt, $text, '', $trail, $prefix );
1746 $this->mOutput
->addLink( $nt );
1750 $s .= $this->makeLinkHolder( $nt, $text, '', $trail, $prefix );
1752 wfProfileOut( $fname );
1757 * Make a link placeholder. The text returned can be later resolved to a real link with
1758 * replaceLinkHolders(). This is done for two reasons: firstly to avoid further
1759 * parsing of interwiki links, and secondly to allow all existence checks and
1760 * article length checks (for stub links) to be bundled into a single query.
1763 function makeLinkHolder( &$nt, $text = '', $query = '', $trail = '', $prefix = '' ) {
1764 wfProfileIn( __METHOD__
);
1765 if ( ! is_object($nt) ) {
1767 $retVal = "<!-- ERROR -->{$prefix}{$text}{$trail}";
1769 # Separate the link trail from the rest of the link
1770 list( $inside, $trail ) = Linker
::splitTrail( $trail );
1772 if ( $nt->isExternal() ) {
1773 $nr = array_push( $this->mInterwikiLinkHolders
['texts'], $prefix.$text.$inside );
1774 $this->mInterwikiLinkHolders
['titles'][] = $nt;
1775 $retVal = '<!--IWLINK '. ($nr-1) ."-->{$trail}";
1777 $nr = array_push( $this->mLinkHolders
['namespaces'], $nt->getNamespace() );
1778 $this->mLinkHolders
['dbkeys'][] = $nt->getDBkey();
1779 $this->mLinkHolders
['queries'][] = $query;
1780 $this->mLinkHolders
['texts'][] = $prefix.$text.$inside;
1781 $this->mLinkHolders
['titles'][] = $nt;
1783 $retVal = '<!--LINK '. ($nr-1) ."-->{$trail}";
1786 wfProfileOut( __METHOD__
);
1791 * Render a forced-blue link inline; protect against double expansion of
1792 * URLs if we're in a mode that prepends full URL prefixes to internal links.
1793 * Since this little disaster has to split off the trail text to avoid
1794 * breaking URLs in the following text without breaking trails on the
1795 * wiki links, it's been made into a horrible function.
1798 * @param string $text
1799 * @param string $query
1800 * @param string $trail
1801 * @param string $prefix
1802 * @return string HTML-wikitext mix oh yuck
1804 function makeKnownLinkHolder( $nt, $text = '', $query = '', $trail = '', $prefix = '' ) {
1805 list( $inside, $trail ) = Linker
::splitTrail( $trail );
1806 $sk = $this->mOptions
->getSkin();
1807 $link = $sk->makeKnownLinkObj( $nt, $text, $query, $inside, $prefix );
1808 return $this->armorLinks( $link ) . $trail;
1812 * Insert a NOPARSE hacky thing into any inline links in a chunk that's
1813 * going to go through further parsing steps before inline URL expansion.
1815 * In particular this is important when using action=render, which causes
1816 * full URLs to be included.
1818 * Oh man I hate our multi-layer parser!
1820 * @param string more-or-less HTML
1821 * @return string less-or-more HTML with NOPARSE bits
1823 function armorLinks( $text ) {
1824 return preg_replace( '/\b(' . wfUrlProtocols() . ')/',
1825 "{$this->mUniqPrefix}NOPARSE$1", $text );
1829 * Return true if subpage links should be expanded on this page.
1832 function areSubpagesAllowed() {
1833 # Some namespaces don't allow subpages
1834 global $wgNamespacesWithSubpages;
1835 return !empty($wgNamespacesWithSubpages[$this->mTitle
->getNamespace()]);
1839 * Handle link to subpage if necessary
1840 * @param string $target the source of the link
1841 * @param string &$text the link text, modified as necessary
1842 * @return string the full name of the link
1845 function maybeDoSubpageLink($target, &$text) {
1848 # :Foobar -- override special treatment of prefix (images, language links)
1849 # /Foobar -- convert to CurrentPage/Foobar
1850 # /Foobar/ -- convert to CurrentPage/Foobar, strip the initial / from text
1851 # ../ -- convert to CurrentPage, from CurrentPage/CurrentSubPage
1852 # ../Foobar -- convert to CurrentPage/Foobar, from CurrentPage/CurrentSubPage
1854 $fname = 'Parser::maybeDoSubpageLink';
1855 wfProfileIn( $fname );
1856 $ret = $target; # default return value is no change
1858 # Some namespaces don't allow subpages,
1859 # so only perform processing if subpages are allowed
1860 if( $this->areSubpagesAllowed() ) {
1861 $hash = strpos( $target, '#' );
1862 if( $hash !== false ) {
1863 $suffix = substr( $target, $hash );
1864 $target = substr( $target, 0, $hash );
1869 $target = trim( $target );
1870 # Look at the first character
1871 if( $target != '' && $target{0} == '/' ) {
1872 # / at end means we don't want the slash to be shown
1874 $trailingSlashes = preg_match_all( '%(/+)$%', $target, $m );
1875 if( $trailingSlashes ) {
1876 $noslash = $target = substr( $target, 1, -strlen($m[0][0]) );
1878 $noslash = substr( $target, 1 );
1881 $ret = $this->mTitle
->getPrefixedText(). '/' . trim($noslash) . $suffix;
1882 if( '' === $text ) {
1883 $text = $target . $suffix;
1884 } # this might be changed for ugliness reasons
1886 # check for .. subpage backlinks
1888 $nodotdot = $target;
1889 while( strncmp( $nodotdot, "../", 3 ) == 0 ) {
1891 $nodotdot = substr( $nodotdot, 3 );
1893 if($dotdotcount > 0) {
1894 $exploded = explode( '/', $this->mTitle
->GetPrefixedText() );
1895 if( count( $exploded ) > $dotdotcount ) { # not allowed to go below top level page
1896 $ret = implode( '/', array_slice( $exploded, 0, -$dotdotcount ) );
1897 # / at the end means don't show full path
1898 if( substr( $nodotdot, -1, 1 ) == '/' ) {
1899 $nodotdot = substr( $nodotdot, 0, -1 );
1900 if( '' === $text ) {
1901 $text = $nodotdot . $suffix;
1904 $nodotdot = trim( $nodotdot );
1905 if( $nodotdot != '' ) {
1906 $ret .= '/' . $nodotdot;
1914 wfProfileOut( $fname );
1919 * Used by doBlockLevels()
1922 /* private */ function closeParagraph() {
1924 if ( '' != $this->mLastSection
) {
1925 $result = '</' . $this->mLastSection
. ">\n";
1927 $this->mInPre
= false;
1928 $this->mLastSection
= '';
1931 # getCommon() returns the length of the longest common substring
1932 # of both arguments, starting at the beginning of both.
1934 /* private */ function getCommon( $st1, $st2 ) {
1935 $fl = strlen( $st1 );
1936 $shorter = strlen( $st2 );
1937 if ( $fl < $shorter ) { $shorter = $fl; }
1939 for ( $i = 0; $i < $shorter; ++
$i ) {
1940 if ( $st1{$i} != $st2{$i} ) { break; }
1944 # These next three functions open, continue, and close the list
1945 # element appropriate to the prefix character passed into them.
1947 /* private */ function openList( $char ) {
1948 $result = $this->closeParagraph();
1950 if ( '*' == $char ) { $result .= '<ul><li>'; }
1951 else if ( '#' == $char ) { $result .= '<ol><li>'; }
1952 else if ( ':' == $char ) { $result .= '<dl><dd>'; }
1953 else if ( ';' == $char ) {
1954 $result .= '<dl><dt>';
1955 $this->mDTopen
= true;
1957 else { $result = '<!-- ERR 1 -->'; }
1962 /* private */ function nextItem( $char ) {
1963 if ( '*' == $char ||
'#' == $char ) { return '</li><li>'; }
1964 else if ( ':' == $char ||
';' == $char ) {
1966 if ( $this->mDTopen
) { $close = '</dt>'; }
1967 if ( ';' == $char ) {
1968 $this->mDTopen
= true;
1969 return $close . '<dt>';
1971 $this->mDTopen
= false;
1972 return $close . '<dd>';
1975 return '<!-- ERR 2 -->';
1978 /* private */ function closeList( $char ) {
1979 if ( '*' == $char ) { $text = '</li></ul>'; }
1980 else if ( '#' == $char ) { $text = '</li></ol>'; }
1981 else if ( ':' == $char ) {
1982 if ( $this->mDTopen
) {
1983 $this->mDTopen
= false;
1984 $text = '</dt></dl>';
1986 $text = '</dd></dl>';
1989 else { return '<!-- ERR 3 -->'; }
1995 * Make lists from lines starting with ':', '*', '#', etc.
1998 * @return string the lists rendered as HTML
2000 function doBlockLevels( $text, $linestart ) {
2001 $fname = 'Parser::doBlockLevels';
2002 wfProfileIn( $fname );
2004 # Parsing through the text line by line. The main thing
2005 # happening here is handling of block-level elements p, pre,
2006 # and making lists from lines starting with * # : etc.
2008 $textLines = explode( "\n", $text );
2010 $lastPrefix = $output = '';
2011 $this->mDTopen
= $inBlockElem = false;
2013 $paragraphStack = false;
2015 if ( !$linestart ) {
2016 $output .= array_shift( $textLines );
2018 foreach ( $textLines as $oLine ) {
2019 $lastPrefixLength = strlen( $lastPrefix );
2020 $preCloseMatch = preg_match('/<\\/pre/i', $oLine );
2021 $preOpenMatch = preg_match('/<pre/i', $oLine );
2022 if ( !$this->mInPre
) {
2023 # Multiple prefixes may abut each other for nested lists.
2024 $prefixLength = strspn( $oLine, '*#:;' );
2025 $pref = substr( $oLine, 0, $prefixLength );
2028 $pref2 = str_replace( ';', ':', $pref );
2029 $t = substr( $oLine, $prefixLength );
2030 $this->mInPre
= !empty($preOpenMatch);
2032 # Don't interpret any other prefixes in preformatted text
2034 $pref = $pref2 = '';
2039 if( $prefixLength && 0 == strcmp( $lastPrefix, $pref2 ) ) {
2040 # Same as the last item, so no need to deal with nesting or opening stuff
2041 $output .= $this->nextItem( substr( $pref, -1 ) );
2042 $paragraphStack = false;
2044 if ( substr( $pref, -1 ) == ';') {
2045 # The one nasty exception: definition lists work like this:
2046 # ; title : definition text
2047 # So we check for : in the remainder text to split up the
2048 # title and definition, without b0rking links.
2050 if ($this->findColonNoLinks($t, $term, $t2) !== false) {
2052 $output .= $term . $this->nextItem( ':' );
2055 } elseif( $prefixLength ||
$lastPrefixLength ) {
2056 # Either open or close a level...
2057 $commonPrefixLength = $this->getCommon( $pref, $lastPrefix );
2058 $paragraphStack = false;
2060 while( $commonPrefixLength < $lastPrefixLength ) {
2061 $output .= $this->closeList( $lastPrefix{$lastPrefixLength-1} );
2062 --$lastPrefixLength;
2064 if ( $prefixLength <= $commonPrefixLength && $commonPrefixLength > 0 ) {
2065 $output .= $this->nextItem( $pref{$commonPrefixLength-1} );
2067 while ( $prefixLength > $commonPrefixLength ) {
2068 $char = substr( $pref, $commonPrefixLength, 1 );
2069 $output .= $this->openList( $char );
2071 if ( ';' == $char ) {
2072 # FIXME: This is dupe of code above
2073 if ($this->findColonNoLinks($t, $term, $t2) !== false) {
2075 $output .= $term . $this->nextItem( ':' );
2078 ++
$commonPrefixLength;
2080 $lastPrefix = $pref2;
2082 if( 0 == $prefixLength ) {
2083 wfProfileIn( "$fname-paragraph" );
2084 # No prefix (not in list)--go to paragraph mode
2085 // XXX: use a stack for nestable elements like span, table and div
2086 $openmatch = preg_match('/(?:<table|<blockquote|<h1|<h2|<h3|<h4|<h5|<h6|<pre|<tr|<p|<ul|<ol|<li|<\\/tr|<\\/td|<\\/th)/iS', $t );
2087 $closematch = preg_match(
2088 '/(?:<\\/table|<\\/blockquote|<\\/h1|<\\/h2|<\\/h3|<\\/h4|<\\/h5|<\\/h6|'.
2089 '<td|<th|<\\/?div|<hr|<\\/pre|<\\/p|'.$this->mUniqPrefix
.'-pre|<\\/li|<\\/ul|<\\/ol|<\\/?center)/iS', $t );
2090 if ( $openmatch or $closematch ) {
2091 $paragraphStack = false;
2092 #Â TODO bug 5718: paragraph closed
2093 $output .= $this->closeParagraph();
2094 if ( $preOpenMatch and !$preCloseMatch ) {
2095 $this->mInPre
= true;
2097 if ( $closematch ) {
2098 $inBlockElem = false;
2100 $inBlockElem = true;
2102 } else if ( !$inBlockElem && !$this->mInPre
) {
2103 if ( ' ' == $t{0} and ( $this->mLastSection
== 'pre' or trim($t) != '' ) ) {
2105 if ($this->mLastSection
!= 'pre') {
2106 $paragraphStack = false;
2107 $output .= $this->closeParagraph().'<pre>';
2108 $this->mLastSection
= 'pre';
2110 $t = substr( $t, 1 );
2113 if ( '' == trim($t) ) {
2114 if ( $paragraphStack ) {
2115 $output .= $paragraphStack.'<br />';
2116 $paragraphStack = false;
2117 $this->mLastSection
= 'p';
2119 if ($this->mLastSection
!= 'p' ) {
2120 $output .= $this->closeParagraph();
2121 $this->mLastSection
= '';
2122 $paragraphStack = '<p>';
2124 $paragraphStack = '</p><p>';
2128 if ( $paragraphStack ) {
2129 $output .= $paragraphStack;
2130 $paragraphStack = false;
2131 $this->mLastSection
= 'p';
2132 } else if ($this->mLastSection
!= 'p') {
2133 $output .= $this->closeParagraph().'<p>';
2134 $this->mLastSection
= 'p';
2139 wfProfileOut( "$fname-paragraph" );
2141 // somewhere above we forget to get out of pre block (bug 785)
2142 if($preCloseMatch && $this->mInPre
) {
2143 $this->mInPre
= false;
2145 if ($paragraphStack === false) {
2149 while ( $prefixLength ) {
2150 $output .= $this->closeList( $pref2{$prefixLength-1} );
2153 if ( '' != $this->mLastSection
) {
2154 $output .= '</' . $this->mLastSection
. '>';
2155 $this->mLastSection
= '';
2158 wfProfileOut( $fname );
2163 * Split up a string on ':', ignoring any occurences inside tags
2164 * to prevent illegal overlapping.
2165 * @param string $str the string to split
2166 * @param string &$before set to everything before the ':'
2167 * @param string &$after set to everything after the ':'
2168 * return string the position of the ':', or false if none found
2170 function findColonNoLinks($str, &$before, &$after) {
2171 $fname = 'Parser::findColonNoLinks';
2172 wfProfileIn( $fname );
2174 $pos = strpos( $str, ':' );
2175 if( $pos === false ) {
2177 wfProfileOut( $fname );
2181 $lt = strpos( $str, '<' );
2182 if( $lt === false ||
$lt > $pos ) {
2183 // Easy; no tag nesting to worry about
2184 $before = substr( $str, 0, $pos );
2185 $after = substr( $str, $pos+
1 );
2186 wfProfileOut( $fname );
2190 // Ugly state machine to walk through avoiding tags.
2191 $state = self
::COLON_STATE_TEXT
;
2193 $len = strlen( $str );
2194 for( $i = 0; $i < $len; $i++
) {
2198 // (Using the number is a performance hack for common cases)
2199 case 0: // self::COLON_STATE_TEXT:
2202 // Could be either a <start> tag or an </end> tag
2203 $state = self
::COLON_STATE_TAGSTART
;
2208 $before = substr( $str, 0, $i );
2209 $after = substr( $str, $i +
1 );
2210 wfProfileOut( $fname );
2213 // Embedded in a tag; don't break it.
2216 // Skip ahead looking for something interesting
2217 $colon = strpos( $str, ':', $i );
2218 if( $colon === false ) {
2219 // Nothing else interesting
2220 wfProfileOut( $fname );
2223 $lt = strpos( $str, '<', $i );
2224 if( $stack === 0 ) {
2225 if( $lt === false ||
$colon < $lt ) {
2227 $before = substr( $str, 0, $colon );
2228 $after = substr( $str, $colon +
1 );
2229 wfProfileOut( $fname );
2233 if( $lt === false ) {
2234 // Nothing else interesting to find; abort!
2235 // We're nested, but there's no close tags left. Abort!
2238 // Skip ahead to next tag start
2240 $state = self
::COLON_STATE_TAGSTART
;
2243 case 1: // self::COLON_STATE_TAG:
2248 $state = self
::COLON_STATE_TEXT
;
2251 // Slash may be followed by >?
2252 $state = self
::COLON_STATE_TAGSLASH
;
2258 case 2: // self::COLON_STATE_TAGSTART:
2261 $state = self
::COLON_STATE_CLOSETAG
;
2264 $state = self
::COLON_STATE_COMMENT
;
2267 // Illegal early close? This shouldn't happen D:
2268 $state = self
::COLON_STATE_TEXT
;
2271 $state = self
::COLON_STATE_TAG
;
2274 case 3: // self::COLON_STATE_CLOSETAG:
2279 wfDebug( "Invalid input in $fname; too many close tags\n" );
2280 wfProfileOut( $fname );
2283 $state = self
::COLON_STATE_TEXT
;
2286 case self
::COLON_STATE_TAGSLASH
:
2288 // Yes, a self-closed tag <blah/>
2289 $state = self
::COLON_STATE_TEXT
;
2291 // Probably we're jumping the gun, and this is an attribute
2292 $state = self
::COLON_STATE_TAG
;
2295 case 5: // self::COLON_STATE_COMMENT:
2297 $state = self
::COLON_STATE_COMMENTDASH
;
2300 case self
::COLON_STATE_COMMENTDASH
:
2302 $state = self
::COLON_STATE_COMMENTDASHDASH
;
2304 $state = self
::COLON_STATE_COMMENT
;
2307 case self
::COLON_STATE_COMMENTDASHDASH
:
2309 $state = self
::COLON_STATE_TEXT
;
2311 $state = self
::COLON_STATE_COMMENT
;
2315 throw new MWException( "State machine error in $fname" );
2319 wfDebug( "Invalid input in $fname; not enough close tags (stack $stack, state $state)\n" );
2322 wfProfileOut( $fname );
2327 * Return value of a magic variable (like PAGENAME)
2331 function getVariableValue( $index ) {
2332 global $wgContLang, $wgSitename, $wgServer, $wgServerName, $wgScriptPath;
2335 * Some of these require message or data lookups and can be
2336 * expensive to check many times.
2338 static $varCache = array();
2339 if ( wfRunHooks( 'ParserGetVariableValueVarCache', array( &$this, &$varCache ) ) ) {
2340 if ( isset( $varCache[$index] ) ) {
2341 return $varCache[$index];
2346 wfRunHooks( 'ParserGetVariableValueTs', array( &$this, &$ts ) );
2349 global $wgLocaltimezone;
2350 if ( isset( $wgLocaltimezone ) ) {
2351 $oldtz = getenv( 'TZ' );
2352 putenv( 'TZ='.$wgLocaltimezone );
2355 wfSuppressWarnings(); // E_STRICT system time bitching
2356 $localTimestamp = date( 'YmdHis', $ts );
2357 $localMonth = date( 'm', $ts );
2358 $localMonthName = date( 'n', $ts );
2359 $localDay = date( 'j', $ts );
2360 $localDay2 = date( 'd', $ts );
2361 $localDayOfWeek = date( 'w', $ts );
2362 $localWeek = date( 'W', $ts );
2363 $localYear = date( 'Y', $ts );
2364 $localHour = date( 'H', $ts );
2365 if ( isset( $wgLocaltimezone ) ) {
2366 putenv( 'TZ='.$oldtz );
2368 wfRestoreWarnings();
2371 case 'currentmonth':
2372 return $varCache[$index] = $wgContLang->formatNum( gmdate( 'm', $ts ) );
2373 case 'currentmonthname':
2374 return $varCache[$index] = $wgContLang->getMonthName( gmdate( 'n', $ts ) );
2375 case 'currentmonthnamegen':
2376 return $varCache[$index] = $wgContLang->getMonthNameGen( gmdate( 'n', $ts ) );
2377 case 'currentmonthabbrev':
2378 return $varCache[$index] = $wgContLang->getMonthAbbreviation( gmdate( 'n', $ts ) );
2380 return $varCache[$index] = $wgContLang->formatNum( gmdate( 'j', $ts ) );
2382 return $varCache[$index] = $wgContLang->formatNum( gmdate( 'd', $ts ) );
2384 return $varCache[$index] = $wgContLang->formatNum( $localMonth );
2385 case 'localmonthname':
2386 return $varCache[$index] = $wgContLang->getMonthName( $localMonthName );
2387 case 'localmonthnamegen':
2388 return $varCache[$index] = $wgContLang->getMonthNameGen( $localMonthName );
2389 case 'localmonthabbrev':
2390 return $varCache[$index] = $wgContLang->getMonthAbbreviation( $localMonthName );
2392 return $varCache[$index] = $wgContLang->formatNum( $localDay );
2394 return $varCache[$index] = $wgContLang->formatNum( $localDay2 );
2396 return wfEscapeWikiText( $this->mTitle
->getText() );
2398 return $this->mTitle
->getPartialURL();
2399 case 'fullpagename':
2400 return wfEscapeWikiText( $this->mTitle
->getPrefixedText() );
2401 case 'fullpagenamee':
2402 return $this->mTitle
->getPrefixedURL();
2404 return wfEscapeWikiText( $this->mTitle
->getSubpageText() );
2405 case 'subpagenamee':
2406 return $this->mTitle
->getSubpageUrlForm();
2407 case 'basepagename':
2408 return wfEscapeWikiText( $this->mTitle
->getBaseText() );
2409 case 'basepagenamee':
2410 return wfUrlEncode( str_replace( ' ', '_', $this->mTitle
->getBaseText() ) );
2411 case 'talkpagename':
2412 if( $this->mTitle
->canTalk() ) {
2413 $talkPage = $this->mTitle
->getTalkPage();
2414 return wfEscapeWikiText( $talkPage->getPrefixedText() );
2418 case 'talkpagenamee':
2419 if( $this->mTitle
->canTalk() ) {
2420 $talkPage = $this->mTitle
->getTalkPage();
2421 return $talkPage->getPrefixedUrl();
2425 case 'subjectpagename':
2426 $subjPage = $this->mTitle
->getSubjectPage();
2427 return wfEscapeWikiText( $subjPage->getPrefixedText() );
2428 case 'subjectpagenamee':
2429 $subjPage = $this->mTitle
->getSubjectPage();
2430 return $subjPage->getPrefixedUrl();
2432 // Let the edit saving system know we should parse the page
2433 // *after* a revision ID has been assigned.
2434 $this->mOutput
->setFlag( 'vary-revision' );
2435 wfDebug( __METHOD__
. ": {{REVISIONID}} used, setting vary-revision...\n" );
2436 return $this->mRevisionId
;
2438 return intval( substr( $this->getRevisionTimestamp(), 6, 2 ) );
2439 case 'revisionday2':
2440 return substr( $this->getRevisionTimestamp(), 6, 2 );
2441 case 'revisionmonth':
2442 return intval( substr( $this->getRevisionTimestamp(), 4, 2 ) );
2443 case 'revisionyear':
2444 return substr( $this->getRevisionTimestamp(), 0, 4 );
2445 case 'revisiontimestamp':
2446 return $this->getRevisionTimestamp();
2448 return str_replace('_',' ',$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 $varCache[$index] = $wgContLang->getWeekdayName( gmdate( 'w', $ts ) +
1 );
2462 return $varCache[$index] = $wgContLang->formatNum( gmdate( 'Y', $ts ), true );
2464 return $varCache[$index] = $wgContLang->time( wfTimestamp( TS_MW
, $ts ), false, false );
2466 return $varCache[$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 $varCache[$index] = $wgContLang->formatNum( (int)gmdate( 'W', $ts ) );
2472 return $varCache[$index] = $wgContLang->formatNum( gmdate( 'w', $ts ) );
2473 case 'localdayname':
2474 return $varCache[$index] = $wgContLang->getWeekdayName( $localDayOfWeek +
1 );
2476 return $varCache[$index] = $wgContLang->formatNum( $localYear, true );
2478 return $varCache[$index] = $wgContLang->time( $localTimestamp, false, false );
2480 return $varCache[$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 $varCache[$index] = $wgContLang->formatNum( (int)$localWeek );
2486 return $varCache[$index] = $wgContLang->formatNum( $localDayOfWeek );
2487 case 'numberofarticles':
2488 return $varCache[$index] = $wgContLang->formatNum( SiteStats
::articles() );
2489 case 'numberoffiles':
2490 return $varCache[$index] = $wgContLang->formatNum( SiteStats
::images() );
2491 case 'numberofusers':
2492 return $varCache[$index] = $wgContLang->formatNum( SiteStats
::users() );
2493 case 'numberofpages':
2494 return $varCache[$index] = $wgContLang->formatNum( SiteStats
::pages() );
2495 case 'numberofadmins':
2496 return $varCache[$index] = $wgContLang->formatNum( SiteStats
::admins() );
2497 case 'numberofedits':
2498 return $varCache[$index] = $wgContLang->formatNum( SiteStats
::edits() );
2499 case 'currenttimestamp':
2500 return $varCache[$index] = wfTimestampNow();
2501 case 'localtimestamp':
2502 return $varCache[$index] = $localTimestamp;
2503 case 'currentversion':
2504 return $varCache[$index] = SpecialVersion
::getVersion();
2510 return $wgServerName;
2512 return $wgScriptPath;
2513 case 'directionmark':
2514 return $wgContLang->getDirMark();
2515 case 'contentlanguage':
2516 global $wgContLanguageCode;
2517 return $wgContLanguageCode;
2520 if ( wfRunHooks( 'ParserGetVariableValueSwitch', array( &$this, &$varCache, &$index, &$ret ) ) )
2528 * initialise the magic variables (like CURRENTMONTHNAME)
2532 function initialiseVariables() {
2533 $fname = 'Parser::initialiseVariables';
2534 wfProfileIn( $fname );
2535 $variableIDs = MagicWord
::getVariableIDs();
2537 $this->mVariables
= new MagicWordArray( $variableIDs );
2538 wfProfileOut( $fname );
2542 * Preprocess some wikitext and return the document tree.
2543 * This is the ghost of replace_variables().
2545 * @param string $text The text to parse
2546 * @param integer flags Bitwise combination of:
2547 * self::PTD_FOR_INCLUSION Handle <noinclude>/<includeonly> as if the text is being
2548 * included. Default is to assume a direct page view.
2550 * The generated DOM tree must depend only on the input text, the flags, and $this->ot['msg'].
2551 * The DOM tree must be the same in OT_HTML and OT_WIKI mode, to avoid a regression of bug 4899.
2553 * Any flag added to the $flags parameter here, or any other parameter liable to cause a
2554 * change in the DOM tree for a given text, must be passed through the section identifier
2555 * in the section edit link and thus back to extractSections().
2557 * The output of this function is currently only cached in process memory, but a persistent
2558 * cache may be implemented at a later date which takes further advantage of these strict
2559 * dependency requirements.
2563 function preprocessToDom ( $text, $flags = 0 ) {
2564 wfProfileIn( __METHOD__
);
2565 wfProfileIn( __METHOD__
.'-makexml' );
2567 static $msgRules, $normalRules, $inclusionSupertags, $nonInclusionSupertags;
2580 'names' => array( 2 => null ),
2585 $normalRules = array(
2597 'names' => array( 2 => null ),
2603 if ( $this->ot
['msg'] ) {
2606 $rules = $normalRules;
2608 $forInclusion = $flags & self
::PTD_FOR_INCLUSION
;
2610 $xmlishElements = $this->getStripList();
2611 $enableOnlyinclude = false;
2612 if ( $forInclusion ) {
2613 $ignoredTags = array( 'includeonly', '/includeonly' );
2614 $ignoredElements = array( 'noinclude' );
2615 $xmlishElements[] = 'noinclude';
2616 if ( strpos( $text, '<onlyinclude>' ) !== false && strpos( $text, '</onlyinclude>' ) !== false ) {
2617 $enableOnlyinclude = true;
2620 $ignoredTags = array( 'noinclude', '/noinclude', 'onlyinclude', '/onlyinclude' );
2621 $ignoredElements = array( 'includeonly' );
2622 $xmlishElements[] = 'includeonly';
2624 $xmlishRegex = implode( '|', array_merge( $xmlishElements, $ignoredTags ) );
2626 // Use "A" modifier (anchored) instead of "^", because ^ doesn't work with an offset
2627 $elementsRegex = "~($xmlishRegex)(?:\s|\/>|>)|(!--)~iA";
2629 $stack = array(); # Stack of unclosed parentheses
2630 $stackIndex = -1; # Stack read pointer
2632 $searchBase = implode( '', array_keys( $rules ) ) . '<';
2633 $revText = strrev( $text ); // For fast reverse searches
2635 $i = -1; # Input pointer, starts out pointing to a pseudo-newline before the start
2636 $topAccum = '<root>'; # Top level text accumulator
2637 $accum =& $topAccum; # Current text accumulator
2638 $findEquals = false; # True to find equals signs in arguments
2639 $findHeading = false; # True to look at LF characters for possible headings
2640 $findPipe = false; # True to take notice of pipe characters
2642 $noMoreGT = false; # True if there are no more greater-than (>) signs right of $i
2643 $findOnlyinclude = $enableOnlyinclude; # True to ignore all input up to the next <onlyinclude>
2645 if ( $enableOnlyinclude ) {
2650 if ( $findOnlyinclude ) {
2651 // Ignore all input up to the next <onlyinclude>
2652 $startPos = strpos( $text, '<onlyinclude>', $i );
2653 if ( $startPos === false ) {
2654 // Ignored section runs to the end
2655 $accum .= '<ignore>' . htmlspecialchars( substr( $text, $i ) ) . '</ignore>';
2658 $tagEndPos = $startPos +
strlen( '<onlyinclude>' ); // past-the-end
2659 $accum .= '<ignore>' . htmlspecialchars( substr( $text, $i, $tagEndPos - $i ) ) . '</ignore>';
2661 $findOnlyinclude = false;
2665 $found = 'line-start';
2668 # Find next opening brace, closing brace or pipe
2669 $search = $searchBase;
2670 if ( $stackIndex == -1 ) {
2671 $currentClosing = '';
2672 // Look for headings only at the top stack level
2673 // Among other things, this resolves the ambiguity between =
2674 // for headings and = for template arguments
2677 $currentClosing = $stack[$stackIndex]['close'];
2678 $search .= $currentClosing;
2683 if ( $findEquals ) {
2687 # Output literal section, advance input counter
2688 $literalLength = strcspn( $text, $search, $i );
2689 if ( $literalLength > 0 ) {
2690 $accum .= htmlspecialchars( substr( $text, $i, $literalLength ) );
2691 $i +
= $literalLength;
2693 if ( $i >= strlen( $text ) ) {
2694 if ( $currentClosing == "\n" ) {
2695 // Do a past-the-end run to finish off the heading
2697 $found = 'line-end';
2703 $curChar = $text[$i];
2704 if ( $curChar == '|' ) {
2706 } elseif ( $curChar == '=' ) {
2708 } elseif ( $curChar == '<' ) {
2710 } elseif ( $curChar == "\n" ) {
2711 if ( $stackIndex == -1 ) {
2712 $found = 'line-start';
2714 $found = 'line-end';
2716 } elseif ( $curChar == $currentClosing ) {
2718 } elseif ( isset( $rules[$curChar] ) ) {
2720 $rule = $rules[$curChar];
2722 # Some versions of PHP have a strcspn which stops on null characters
2723 # Ignore and continue
2730 if ( $found == 'angle' ) {
2732 // Handle </onlyinclude>
2733 if ( $enableOnlyinclude && substr( $text, $i, strlen( '</onlyinclude>' ) ) == '</onlyinclude>' ) {
2734 $findOnlyinclude = true;
2738 // Determine element name
2739 if ( !preg_match( $elementsRegex, $text, $matches, 0, $i +
1 ) ) {
2740 // Element name missing or not listed
2746 if ( isset( $matches[2] ) && $matches[2] == '!--' ) {
2747 // To avoid leaving blank lines, when a comment is both preceded
2748 // and followed by a newline (ignoring spaces), trim leading and
2749 // trailing spaces and one of the newlines.
2752 $endPos = strpos( $text, '-->', $i +
4 );
2753 if ( $endPos === false ) {
2754 // Unclosed comment in input, runs to end
2755 $inner = substr( $text, $i );
2756 $accum .= '<comment>' . htmlspecialchars( $inner ) . '</comment>';
2757 $i = strlen( $text );
2759 // Search backwards for leading whitespace
2760 $wsStart = $i ?
( $i - strspn( $revText, ' ', strlen( $text ) - $i - 1 ) ) : 0;
2761 // Search forwards for trailing whitespace
2762 // $wsEnd will be the position of the last space
2763 $wsEnd = $endPos +
2 +
strspn( $text, ' ', $endPos +
3 );
2764 // Eat the line if possible
2765 if ( $wsStart > 0 && substr( $text, $wsStart - 1, 1 ) == "\n"
2766 && substr( $text, $wsEnd +
1, 1 ) == "\n" )
2768 $startPos = $wsStart;
2769 $endPos = $wsEnd +
1;
2771 // No line to eat, just take the comment itself
2775 $inner = substr( $text, $startPos, $endPos - $startPos +
1 );
2776 $accum .= '<comment>' . htmlspecialchars( $inner ) . '</comment>';
2781 $name = $matches[1];
2782 $attrStart = $i +
strlen( $name ) +
1;
2785 $tagEndPos = $noMoreGT ?
false : strpos( $text, '>', $attrStart );
2786 if ( $tagEndPos === false ) {
2787 // Infinite backtrack
2788 // Disable tag search to prevent worst-case O(N^2) performance
2795 // Handle ignored tags
2796 if ( in_array( $name, $ignoredTags ) ) {
2797 $accum .= '<ignore>' . htmlspecialchars( substr( $text, $i, $tagEndPos - $i +
1 ) ) . '</ignore>';
2798 $i = $tagEndPos +
1;
2803 if ( $text[$tagEndPos-1] == '/' ) {
2804 $attrEnd = $tagEndPos - 1;
2806 $i = $tagEndPos +
1;
2809 $attrEnd = $tagEndPos;
2811 if ( preg_match( "/<\/$name\s*>/i", $text, $matches, PREG_OFFSET_CAPTURE
, $tagEndPos +
1 ) ) {
2812 $inner = substr( $text, $tagEndPos +
1, $matches[0][1] - $tagEndPos - 1 );
2813 $i = $matches[0][1] +
strlen( $matches[0][0] );
2814 $close = '<close>' . htmlspecialchars( $matches[0][0] ) . '</close>';
2816 // No end tag -- let it run out to the end of the text.
2817 $inner = substr( $text, $tagEndPos +
1 );
2818 $i = strlen( $text );
2822 // <includeonly> and <noinclude> just become <ignore> tags
2823 if ( in_array( $name, $ignoredElements ) ) {
2824 $accum .= '<ignore>' . htmlspecialchars( substr( $text, $tagStartPos, $i - $tagStartPos ) )
2830 if ( $attrEnd <= $attrStart ) {
2833 $attr = substr( $text, $attrStart, $attrEnd - $attrStart );
2835 $accum .= '<name>' . htmlspecialchars( $name ) . '</name>' .
2836 // Note that the attr element contains the whitespace between name and attribute,
2837 // this is necessary for precise reconstruction during pre-save transform.
2838 '<attr>' . htmlspecialchars( $attr ) . '</attr>';
2839 if ( $inner !== null ) {
2840 $accum .= '<inner>' . htmlspecialchars( $inner ) . '</inner>';
2842 $accum .= $close . '</ext>';
2845 elseif ( $found == 'line-start' ) {
2846 // Is this the start of a heading?
2847 // Line break belongs before the heading element in any case
2851 $count = strspn( $text, '=', $i, 6 );
2856 'parts' => array( str_repeat( '=', $count ) ),
2857 'count' => $count );
2858 $stack[++
$stackIndex] = $piece;
2860 $accum =& $stack[$stackIndex]['parts'][0];
2865 elseif ( $found == 'line-end' ) {
2866 $piece = $stack[$stackIndex];
2867 // A heading must be open, otherwise \n wouldn't have been in the search list
2868 assert( $piece['open'] == "\n" );
2869 assert( $stackIndex == 0 );
2870 // Search back through the input to see if it has a proper close
2871 // Do this using the reversed string since the other solutions (end anchor, etc.) are inefficient
2873 $count = $piece['count'];
2874 if ( preg_match( "/\s*(={{$count}})/A", $revText, $m, 0, strlen( $text ) - $i ) ) {
2875 // Found match, output <h>
2876 $count = min( strlen( $m[1] ), $count );
2877 $element = "<h level=\"$count\" i=\"$headingIndex\">$accum</h>";
2880 // No match, no <h>, just pass down the inner text
2884 // Headings can only occur on the top level, so this is a bit simpler than the
2885 // generic stack unwind operation in the close case
2886 unset( $stack[$stackIndex--] );
2887 $accum =& $topAccum;
2888 $findEquals = false;
2891 // Append the result to the enclosing accumulator
2893 // Note that we do NOT increment the input pointer.
2894 // This is because the closing linebreak could be the opening linebreak of
2895 // another heading. Infinite loops are avoided because the next iteration MUST
2896 // hit the heading open case above, which unconditionally increments the
2900 elseif ( $found == 'open' ) {
2901 # count opening brace characters
2902 $count = strspn( $text, $curChar, $i );
2904 # we need to add to stack only if opening brace count is enough for one of the rules
2905 if ( $count >= $rule['min'] ) {
2906 # Add it to the stack
2909 'close' => $rule['end'],
2911 'parts' => array( '' ),
2913 'lineStart' => ($i > 0 && $text[$i-1] == "\n"),
2917 $stack[$stackIndex] = $piece;
2918 $accum =& $stack[$stackIndex]['parts'][0];
2919 $findEquals = false;
2922 # Add literal brace(s)
2923 $accum .= htmlspecialchars( str_repeat( $curChar, $count ) );
2928 elseif ( $found == 'close' ) {
2929 $piece = $stack[$stackIndex];
2930 # lets check if there are enough characters for closing brace
2931 $maxCount = $piece['count'];
2932 $count = strspn( $text, $curChar, $i, $maxCount );
2934 # check for maximum matching characters (if there are 5 closing
2935 # characters, we will probably need only 3 - depending on the rules)
2937 $rule = $rules[$piece['open']];
2938 if ( $count > $rule['max'] ) {
2939 # The specified maximum exists in the callback array, unless the caller
2941 $matchingCount = $rule['max'];
2943 # Count is less than the maximum
2944 # Skip any gaps in the callback array to find the true largest match
2945 # Need to use array_key_exists not isset because the callback can be null
2946 $matchingCount = $count;
2947 while ( $matchingCount > 0 && !array_key_exists( $matchingCount, $rule['names'] ) ) {
2952 if ($matchingCount <= 0) {
2953 # No matching element found in callback array
2954 # Output a literal closing brace and continue
2955 $accum .= htmlspecialchars( str_repeat( $curChar, $count ) );
2959 $name = $rule['names'][$matchingCount];
2960 if ( $name === null ) {
2961 // No element, just literal text
2962 $element = str_repeat( $piece['open'], $matchingCount ) .
2963 implode( '|', $piece['parts'] ) .
2964 str_repeat( $rule['end'], $matchingCount );
2966 # Create XML element
2967 # Note: $parts is already XML, does not need to be encoded further
2968 $parts = $piece['parts'];
2972 # The invocation is at the start of the line if lineStart is set in
2973 # the stack, and all opening brackets are used up.
2974 if ( $maxCount == $matchingCount && !empty( $piece['lineStart'] ) ) {
2975 $attr = ' lineStart="1"';
2980 $element = "<$name$attr>";
2981 $element .= "<title>$title</title>";
2983 foreach ( $parts as $partIndex => $part ) {
2984 if ( isset( $piece['eqpos'][$partIndex] ) ) {
2985 $eqpos = $piece['eqpos'][$partIndex];
2986 list( $ws1, $argName, $ws2 ) = self
::splitWhitespace( substr( $part, 0, $eqpos ) );
2987 list( $ws3, $argValue, $ws4 ) = self
::splitWhitespace( substr( $part, $eqpos +
1 ) );
2988 $element .= "<part>$ws1<name>$argName</name>$ws2=$ws3<value>$argValue</value>$ws4</part>";
2990 list( $ws1, $value, $ws2 ) = self
::splitWhitespace( $part );
2991 $element .= "<part>$ws1<name index=\"$argIndex\" /><value>$value</value>$ws2</part>";
2995 $element .= "</$name>";
2998 # Advance input pointer
2999 $i +
= $matchingCount;
3002 unset( $stack[$stackIndex--] );
3003 if ( $stackIndex == -1 ) {
3004 $accum =& $topAccum;
3005 $findEquals = false;
3008 $partCount = count( $stack[$stackIndex]['parts'] );
3009 $accum =& $stack[$stackIndex]['parts'][$partCount - 1];
3010 $findPipe = $stack[$stackIndex]['open'] != "\n";
3011 $findEquals = $findPipe && $partCount > 1
3012 && !isset( $stack[$stackIndex]['eqpos'][$partCount - 1] );
3015 # Re-add the old stack element if it still has unmatched opening characters remaining
3016 if ($matchingCount < $piece['count']) {
3017 $piece['parts'] = array( '' );
3018 $piece['count'] -= $matchingCount;
3019 $piece['eqpos'] = array();
3020 # do we still qualify for any callback with remaining count?
3021 $names = $rules[$piece['open']]['names'];
3023 $enclosingAccum =& $accum;
3024 while ( $piece['count'] ) {
3025 if ( array_key_exists( $piece['count'], $names ) ) {
3027 $stack[$stackIndex] = $piece;
3028 $accum =& $stack[$stackIndex]['parts'][0];
3036 $enclosingAccum .= str_repeat( $piece['open'], $skippedBraces );
3039 # Add XML element to the enclosing accumulator
3043 elseif ( $found == 'pipe' ) {
3044 $stack[$stackIndex]['parts'][] = '';
3045 $partsCount = count( $stack[$stackIndex]['parts'] );
3046 $accum =& $stack[$stackIndex]['parts'][$partsCount - 1];
3051 elseif ( $found == 'equals' ) {
3052 $findEquals = false;
3053 $partsCount = count( $stack[$stackIndex]['parts'] );
3054 $stack[$stackIndex]['eqpos'][$partsCount - 1] = strlen( $accum );
3060 # Output any remaining unclosed brackets
3061 foreach ( $stack as $piece ) {
3062 if ( $piece['open'] == "\n" ) {
3063 $topAccum .= $piece['parts'][0];
3065 $topAccum .= str_repeat( $piece['open'], $piece['count'] ) . implode( '|', $piece['parts'] );
3068 $topAccum .= '</root>';
3070 wfProfileOut( __METHOD__
.'-makexml' );
3071 wfProfileIn( __METHOD__
.'-loadXML' );
3072 $dom = new DOMDocument
;
3073 wfSuppressWarnings();
3074 $result = $dom->loadXML( $topAccum );
3075 wfRestoreWarnings();
3077 // Try running the XML through UtfNormal to get rid of invalid characters
3078 $topAccum = UtfNormal
::cleanUp( $topAccum );
3079 $result = $dom->loadXML( $topAccum );
3081 throw new MWException( __METHOD__
.' generated invalid XML' );
3084 wfProfileOut( __METHOD__
.'-loadXML' );
3085 wfProfileOut( __METHOD__
);
3090 * Return a three-element array: leading whitespace, string contents, trailing whitespace
3092 public static function splitWhitespace( $s ) {
3093 $ltrimmed = ltrim( $s );
3094 $w1 = substr( $s, 0, strlen( $s ) - strlen( $ltrimmed ) );
3095 $trimmed = rtrim( $ltrimmed );
3096 $diff = strlen( $ltrimmed ) - strlen( $trimmed );
3098 $w2 = substr( $ltrimmed, -$diff );
3102 return array( $w1, $trimmed, $w2 );
3106 * Replace magic variables, templates, and template arguments
3107 * with the appropriate text. Templates are substituted recursively,
3108 * taking care to avoid infinite loops.
3110 * Note that the substitution depends on value of $mOutputType:
3111 * OT_WIKI: only {{subst:}} templates
3112 * OT_MSG: only magic variables
3113 * OT_HTML: all templates and magic variables
3115 * @param string $tex The text to transform
3116 * @param PPFrame $frame Object describing the arguments passed to the template
3117 * @param bool $argsOnly Only do argument (triple-brace) expansion, not double-brace expansion
3120 function replaceVariables( $text, $frame = false, $argsOnly = false ) {
3121 # Prevent too big inclusions
3122 if( strlen( $text ) > $this->mOptions
->getMaxIncludeSize() ) {
3126 $fname = __METHOD__
;
3127 wfProfileIn( $fname );
3129 if ( $frame === false ) {
3130 $frame = new PPFrame( $this );
3131 } elseif ( !( $frame instanceof PPFrame
) ) {
3132 throw new MWException( __METHOD__
. ' called using the old argument format' );
3135 $dom = $this->preprocessToDom( $text );
3136 $flags = $argsOnly ? PPFrame
::NO_TEMPLATES
: 0;
3137 $text = $frame->expand( $dom, $flags );
3139 wfProfileOut( $fname );
3143 /// Clean up argument array - refactored in 1.9 so parserfunctions can use it, too.
3144 static function createAssocArgs( $args ) {
3145 $assocArgs = array();
3147 foreach( $args as $arg ) {
3148 $eqpos = strpos( $arg, '=' );
3149 if ( $eqpos === false ) {
3150 $assocArgs[$index++
] = $arg;
3152 $name = trim( substr( $arg, 0, $eqpos ) );
3153 $value = trim( substr( $arg, $eqpos+
1 ) );
3154 if ( $value === false ) {
3157 if ( $name !== false ) {
3158 $assocArgs[$name] = $value;
3167 * Return the text of a template, after recursively
3168 * replacing any variables or templates within the template.
3170 * @param array $piece The parts of the template
3171 * $piece['text']: matched text
3172 * $piece['title']: the title, i.e. the part before the |
3173 * $piece['parts']: the parameter array
3174 * @param PPFrame The current frame, contains template arguments
3175 * @return string the text of the template
3178 function braceSubstitution( $piece, $frame ) {
3179 global $wgContLang, $wgLang, $wgAllowDisplayTitle, $wgNonincludableNamespaces;
3180 $fname = __METHOD__
;
3181 wfProfileIn( $fname );
3182 wfProfileIn( __METHOD__
.'-setup' );
3185 $found = false; # $text has been filled
3186 $nowiki = false; # wiki markup in $text should be escaped
3187 $isHTML = false; # $text is HTML, armour it against wikitext transformation
3188 $forceRawInterwiki = false; # Force interwiki transclusion to be done in raw mode not rendered
3189 $isDOM = false; # $text is a DOM node needing expansion
3191 # Title object, where $text came from
3194 # $part1 is the bit before the first |, and must contain only title characters.
3195 # Various prefixes will be stripped from it later.
3196 $titleWithSpaces = $frame->expand( $piece['title'] );
3197 $part1 = trim( $titleWithSpaces );
3200 # Original title text preserved for various purposes
3201 $originalTitle = $part1;
3203 # $args is a list of argument nodes, starting from index 0, not including $part1
3204 $args = (null == $piece['parts']) ?
array() : $piece['parts'];
3205 wfProfileOut( __METHOD__
.'-setup' );
3208 wfProfileIn( __METHOD__
.'-modifiers' );
3210 $mwSubst =& MagicWord
::get( 'subst' );
3211 if ( $mwSubst->matchStartAndRemove( $part1 ) xor $this->ot
['wiki'] ) {
3212 # One of two possibilities is true:
3213 # 1) Found SUBST but not in the PST phase
3214 # 2) Didn't find SUBST and in the PST phase
3215 # In either case, return without further processing
3216 $text = '{{' . $frame->implode( '|', $titleWithSpaces, $args ) . '}}';
3222 if ( !$found && $args->length
== 0 ) {
3223 $id = $this->mVariables
->matchStartToEnd( $part1 );
3224 if ( $id !== false ) {
3225 $text = $this->getVariableValue( $id );
3226 $this->mOutput
->mContainsOldMagic
= true;
3231 # MSG, MSGNW and RAW
3234 $mwMsgnw =& MagicWord
::get( 'msgnw' );
3235 if ( $mwMsgnw->matchStartAndRemove( $part1 ) ) {
3238 # Remove obsolete MSG:
3239 $mwMsg =& MagicWord
::get( 'msg' );
3240 $mwMsg->matchStartAndRemove( $part1 );
3244 $mwRaw =& MagicWord
::get( 'raw' );
3245 if ( $mwRaw->matchStartAndRemove( $part1 ) ) {
3246 $forceRawInterwiki = true;
3249 wfProfileOut( __METHOD__
.'-modifiers' );
3251 # Save path level before recursing into functions & templates.
3252 $lastPathLevel = $this->mTemplatePath
;
3256 wfProfileIn( __METHOD__
. '-pfunc' );
3258 $colonPos = strpos( $part1, ':' );
3259 if ( $colonPos !== false ) {
3260 # Case sensitive functions
3261 $function = substr( $part1, 0, $colonPos );
3262 if ( isset( $this->mFunctionSynonyms
[1][$function] ) ) {
3263 $function = $this->mFunctionSynonyms
[1][$function];
3265 # Case insensitive functions
3266 $function = strtolower( $function );
3267 if ( isset( $this->mFunctionSynonyms
[0][$function] ) ) {
3268 $function = $this->mFunctionSynonyms
[0][$function];
3274 list( $callback, $flags ) = $this->mFunctionHooks
[$function];
3275 $initialArgs = array( &$this );
3276 $funcArgs = array( trim( substr( $part1, $colonPos +
1 ) ) );
3277 if ( $flags & SFH_OBJECT_ARGS
) {
3278 # Add a frame parameter, and pass the arguments as an array
3279 $allArgs = $initialArgs;
3280 $allArgs[] = $frame;
3281 foreach ( $args as $arg ) {
3284 $allArgs[] = $funcArgs;
3286 # Convert arguments to plain text
3287 foreach ( $args as $arg ) {
3288 $funcArgs[] = trim( $frame->expand( $arg ) );
3290 $allArgs = array_merge( $initialArgs, $funcArgs );
3293 $result = call_user_func_array( $callback, $allArgs );
3296 if ( is_array( $result ) ) {
3297 if ( isset( $result[0] ) ) {
3299 unset( $result[0] );
3302 // Extract flags into the local scope
3303 // This allows callers to set flags such as nowiki, found, etc.
3310 wfProfileOut( __METHOD__
. '-pfunc' );
3313 # Finish mangling title and then check for loops.
3314 # Set $title to a Title object and $titleText to the PDBK
3317 # Split the title into page and subpage
3319 $part1 = $this->maybeDoSubpageLink( $part1, $subpage );
3320 if ($subpage !== '') {
3321 $ns = $this->mTitle
->getNamespace();
3323 $title = Title
::newFromText( $part1, $ns );
3325 $titleText = $title->getPrefixedText();
3326 # Check for language variants if the template is not found
3327 if($wgContLang->hasVariants() && $title->getArticleID() == 0){
3328 $wgContLang->findVariantLink($part1, $title);
3330 # Do infinite loop check
3331 if ( isset( $this->mTemplatePath
[$titleText] ) ) {
3333 $text = "[[$part1]]" . $this->insertStripItem( '<!-- WARNING: template loop detected -->' );
3334 wfDebug( __METHOD__
.": template loop broken at '$titleText'\n" );
3339 # Load from database
3340 if ( !$found && $title ) {
3341 wfProfileIn( __METHOD__
. '-loadtpl' );
3342 if ( !$title->isExternal() ) {
3343 if ( $title->getNamespace() == NS_SPECIAL
&& $this->mOptions
->getAllowSpecialInclusion() && $this->ot
['html'] ) {
3344 $text = SpecialPage
::capturePath( $title );
3345 if ( is_string( $text ) ) {
3348 $this->disableCache();
3350 } else if ( $wgNonincludableNamespaces && in_array( $title->getNamespace(), $wgNonincludableNamespaces ) ) {
3351 $found = false; //access denied
3352 wfDebug( "$fname: template inclusion denied for " . $title->getPrefixedDBkey() );
3354 list( $text, $title ) = $this->getTemplateDom( $title );
3355 if ( $text !== false ) {
3361 # If the title is valid but undisplayable, make a link to it
3362 if ( !$found && ( $this->ot
['html'] ||
$this->ot
['pre'] ) ) {
3363 $text = "[[:$titleText]]";
3366 } elseif ( $title->isTrans() ) {
3367 // Interwiki transclusion
3368 if ( $this->ot
['html'] && !$forceRawInterwiki ) {
3369 $text = $this->interwikiTransclude( $title, 'render' );
3372 $text = $this->interwikiTransclude( $title, 'raw' );
3373 // Preprocess it like a template
3374 $text = $this->preprocessToDom( $text, self
::PTD_FOR_INCLUSION
);
3379 wfProfileOut( __METHOD__
. '-loadtpl' );
3382 # If we haven't found text to substitute by now, we're done
3383 # Recover the source wikitext and return it
3385 $text = '{{' . $frame->implode( '|', $titleWithSpaces, $args ) . '}}';
3386 # Prune lower levels off the recursion check path
3387 $this->mTemplatePath
= $lastPathLevel;
3388 wfProfileOut( $fname );
3392 # Expand DOM-style return values in a child frame
3394 # Clean up argument array
3395 $newFrame = $frame->newChild( $args, $title );
3396 # Add a new element to the templace recursion path
3397 $this->mTemplatePath
[$titleText] = 1;
3399 if ( $titleText !== false && count( $newFrame->args
) == 0 ) {
3400 # Expansion is eligible for the empty-frame cache
3401 if ( isset( $this->mTplExpandCache
[$titleText] ) ) {
3402 $text = $this->mTplExpandCache
[$titleText];
3404 $text = $newFrame->expand( $text );
3405 $this->mTplExpandCache
[$titleText] = $text;
3408 $text = $newFrame->expand( $text );
3412 # Replace raw HTML by a placeholder
3413 # Add a blank line preceding, to prevent it from mucking up
3414 # immediately preceding headings
3416 $text = "\n\n" . $this->insertStripItem( $text );
3418 # Escape nowiki-style return values
3419 elseif ( $nowiki && ( $this->ot
['html'] ||
$this->ot
['pre'] ) ) {
3420 $text = wfEscapeWikiText( $text );
3422 # Bug 529: if the template begins with a table or block-level
3423 # element, it should be treated as beginning a new line.
3424 # This behaviour is somewhat controversial.
3425 elseif ( !$piece['lineStart'] && preg_match('/^(?:{\\||:|;|#|\*)/', $text)) /*}*/{
3426 $text = "\n" . $text;
3429 # Prune lower levels off the recursion check path
3430 $this->mTemplatePath
= $lastPathLevel;
3432 if ( !$this->incrementIncludeSize( 'post-expand', strlen( $text ) ) ) {
3433 # Error, oversize inclusion
3434 $text = "[[$originalTitle]]" .
3435 $this->insertStripItem( '<!-- WARNING: template omitted, post-expand include size too large -->' );
3438 wfProfileOut( $fname );
3443 * Get the semi-parsed DOM representation of a template with a given title,
3444 * and its redirect destination title. Cached.
3446 function getTemplateDom( $title ) {
3447 $cacheTitle = $title;
3448 $titleText = $title->getPrefixedDBkey();
3450 if ( isset( $this->mTplRedirCache
[$titleText] ) ) {
3451 list( $ns, $dbk ) = $this->mTplRedirCache
[$titleText];
3452 $title = Title
::makeTitle( $ns, $dbk );
3453 $titleText = $title->getPrefixedDBkey();
3455 if ( isset( $this->mTplDomCache
[$titleText] ) ) {
3456 return array( $this->mTplDomCache
[$titleText], $title );
3459 // Cache miss, go to the database
3460 list( $text, $title ) = $this->fetchTemplateAndTitle( $title );
3462 if ( $text === false ) {
3463 $this->mTplDomCache
[$titleText] = false;
3464 return array( false, $title );
3467 $dom = $this->preprocessToDom( $text, self
::PTD_FOR_INCLUSION
);
3468 $this->mTplDomCache
[ $titleText ] = $dom;
3470 if (! $title->equals($cacheTitle)) {
3471 $this->mTplRedirCache
[$cacheTitle->getPrefixedDBkey()] =
3472 array( $title->getNamespace(),$cdb = $title->getDBkey() );
3475 return array( $dom, $title );
3479 * Fetch the unparsed text of a template and register a reference to it.
3481 function fetchTemplateAndTitle( $title ) {
3482 $templateCb = $this->mOptions
->getTemplateCallback();
3483 $stuff = call_user_func( $templateCb, $title );
3484 $text = $stuff['text'];
3485 $finalTitle = isset( $stuff['finalTitle'] ) ?
$stuff['finalTitle'] : $title;
3486 if ( isset( $stuff['deps'] ) ) {
3487 foreach ( $stuff['deps'] as $dep ) {
3488 $this->mOutput
->addTemplate( $dep['title'], $dep['page_id'], $dep['rev_id'] );
3491 return array($text,$finalTitle);
3494 function fetchTemplate( $title ) {
3495 $rv = $this->fetchTemplateAndTitle($title);
3500 * Static function to get a template
3501 * Can be overridden via ParserOptions::setTemplateCallback().
3503 static function statelessFetchTemplate( $title ) {
3504 $text = $skip = false;
3505 $finalTitle = $title;
3508 // Loop to fetch the article, with up to 1 redirect
3509 for ( $i = 0; $i < 2 && is_object( $title ); $i++
) {
3510 # Give extensions a chance to select the revision instead
3511 $id = false; // Assume current
3512 wfRunHooks( 'BeforeParserFetchTemplateAndtitle', array( false, &$title, &$skip, &$id ) );
3518 'page_id' => $title->getArticleID(),
3522 $rev = $id ? Revision
::newFromId( $id ) : Revision
::newFromTitle( $title );
3523 $rev_id = $rev ?
$rev->getId() : 0;
3527 'page_id' => $title->getArticleID(),
3528 'rev_id' => $rev_id );
3531 $text = $rev->getText();
3532 } elseif( $title->getNamespace() == NS_MEDIAWIKI
) {
3534 $message = $wgLang->lcfirst( $title->getText() );
3535 $text = wfMsgForContentNoTrans( $message );
3536 if( wfEmptyMsg( $message, $text ) ) {
3543 if ( $text === false ) {
3547 $finalTitle = $title;
3548 $title = Title
::newFromRedirect( $text );
3552 'finalTitle' => $finalTitle,
3557 * Transclude an interwiki link.
3559 function interwikiTransclude( $title, $action ) {
3560 global $wgEnableScaryTranscluding;
3562 if (!$wgEnableScaryTranscluding)
3563 return wfMsg('scarytranscludedisabled');
3565 $url = $title->getFullUrl( "action=$action" );
3567 if (strlen($url) > 255)
3568 return wfMsg('scarytranscludetoolong');
3569 return $this->fetchScaryTemplateMaybeFromCache($url);
3572 function fetchScaryTemplateMaybeFromCache($url) {
3573 global $wgTranscludeCacheExpiry;
3574 $dbr = wfGetDB(DB_SLAVE
);
3575 $obj = $dbr->selectRow('transcache', array('tc_time', 'tc_contents'),
3576 array('tc_url' => $url));
3578 $time = $obj->tc_time
;
3579 $text = $obj->tc_contents
;
3580 if ($time && time() < $time +
$wgTranscludeCacheExpiry ) {
3585 $text = Http
::get($url);
3587 return wfMsg('scarytranscludefailed', $url);
3589 $dbw = wfGetDB(DB_MASTER
);
3590 $dbw->replace('transcache', array('tc_url'), array(
3592 'tc_time' => time(),
3593 'tc_contents' => $text));
3599 * Triple brace replacement -- used for template arguments
3602 function argSubstitution( $piece, $frame ) {
3603 wfProfileIn( __METHOD__
);
3607 $parts = $piece['parts'];
3608 $argWithSpaces = $frame->expand( $piece['title'] );
3609 $arg = trim( $argWithSpaces );
3611 if ( isset( $frame->args
[$arg] ) ) {
3612 $text = $frame->parent
->expand( $frame->args
[$arg] );
3613 } else if ( ( $this->ot
['html'] ||
$this->ot
['pre'] ) && $parts->length
> 0 ) {
3614 $text = $frame->expand( $parts->item( 0 ) );
3616 if ( !$this->incrementIncludeSize( 'arg', strlen( $text ) ) ) {
3617 $error = '<!-- WARNING: argument omitted, expansion size too large -->';
3620 if ( $text === false ) {
3621 $text = '{{{' . $frame->implode( '|', $argWithSpaces, $parts ) . '}}}';
3623 if ( $error !== false ) {
3627 wfProfileOut( __METHOD__
);
3632 * Return the text to be used for a given extension tag.
3633 * This is the ghost of strip().
3635 * @param array $params Associative array of parameters:
3636 * name DOMNode for the tag name
3637 * attrText DOMNode for unparsed text where tag attributes are thought to be
3638 * inner Contents of extension element
3639 * noClose Original text did not have a close tag
3640 * @param PPFrame $frame
3642 function extensionSubstitution( $params, $frame ) {
3643 global $wgRawHtml, $wgContLang;
3646 $name = $frame->expand( $params['name'] );
3647 $attrText = is_null( $params['attr'] ) ?
null : $frame->expand( $params['attr'] );
3648 $content = is_null( $params['inner'] ) ?
null : $frame->expand( $params['inner'] );
3650 $marker = "{$this->mUniqPrefix}-$name-" . sprintf('%08X', $n++
) . $this->mMarkerSuffix
;
3652 if ( $this->ot
['html'] ) {
3653 $name = strtolower( $name );
3655 $params = Sanitizer
::decodeTagAttributes( $attrText );
3662 throw new MWException( '<html> extension tag encountered unexpectedly' );
3665 $output = Xml
::escapeTagsOnly( $content );
3668 $output = $wgContLang->armourMath(
3669 MathRenderer
::renderMath( $content, $params ) );
3672 $output = $this->renderImageGallery( $content, $params );
3675 if( isset( $this->mTagHooks
[$name] ) ) {
3676 $output = call_user_func_array( $this->mTagHooks
[$name],
3677 array( $content, $params, $this ) );
3679 throw new MWException( "Invalid call hook $name" );
3683 if ( $content === null ) {
3684 $output = "<$name$attrText/>";
3686 $close = is_null( $params['close'] ) ?
'' : $frame->expand( $params['close'] );
3687 $output = "<$name$attrText>$content$close";
3691 if ( $name == 'html' ||
$name == 'nowiki' ) {
3692 $this->mStripState
->nowiki
->setPair( $marker, $output );
3694 $this->mStripState
->general
->setPair( $marker, $output );
3700 * Increment an include size counter
3702 * @param string $type The type of expansion
3703 * @param integer $size The size of the text
3704 * @return boolean False if this inclusion would take it over the maximum, true otherwise
3706 function incrementIncludeSize( $type, $size ) {
3707 if ( $this->mIncludeSizes
[$type] +
$size > $this->mOptions
->getMaxIncludeSize( $type ) ) {
3710 $this->mIncludeSizes
[$type] +
= $size;
3716 * Detect __NOGALLERY__ magic word and set a placeholder
3718 function stripNoGallery( &$text ) {
3719 # if the string __NOGALLERY__ (not case-sensitive) occurs in the HTML,
3721 $mw = MagicWord
::get( 'nogallery' );
3722 $this->mOutput
->mNoGallery
= $mw->matchAndRemove( $text ) ;
3726 * Find the first __TOC__ magic word and set a <!--MWTOC-->
3727 * placeholder that will then be replaced by the real TOC in
3728 * ->formatHeadings, this works because at this points real
3729 * comments will have already been discarded by the sanitizer.
3731 * Any additional __TOC__ magic words left over will be discarded
3732 * as there can only be one TOC on the page.
3734 function stripToc( $text ) {
3735 # if the string __NOTOC__ (not case-sensitive) occurs in the HTML,
3737 $mw = MagicWord
::get( 'notoc' );
3738 if( $mw->matchAndRemove( $text ) ) {
3739 $this->mShowToc
= false;
3742 $mw = MagicWord
::get( 'toc' );
3743 if( $mw->match( $text ) ) {
3744 $this->mShowToc
= true;
3745 $this->mForceTocPosition
= true;
3747 // Set a placeholder. At the end we'll fill it in with the TOC.
3748 $text = $mw->replace( '<!--MWTOC-->', $text, 1 );
3750 // Only keep the first one.
3751 $text = $mw->replace( '', $text );
3757 * This function accomplishes several tasks:
3758 * 1) Auto-number headings if that option is enabled
3759 * 2) Add an [edit] link to sections for users who have enabled the option and can edit the page
3760 * 3) Add a Table of contents on the top for users who have enabled the option
3761 * 4) Auto-anchor headings
3763 * It loops through all headlines, collects the necessary data, then splits up the
3764 * string and re-inserts the newly formatted headlines.
3766 * @param string $text
3767 * @param boolean $isMain
3770 function formatHeadings( $text, $isMain=true ) {
3771 global $wgMaxTocLevel, $wgContLang;
3773 $doNumberHeadings = $this->mOptions
->getNumberHeadings();
3774 if( !$this->mTitle
->quickUserCan( 'edit' ) ) {
3777 $showEditLink = $this->mOptions
->getEditSection();
3780 # Inhibit editsection links if requested in the page
3781 $esw =& MagicWord
::get( 'noeditsection' );
3782 if( $esw->matchAndRemove( $text ) ) {
3786 # Get all headlines for numbering them and adding funky stuff like [edit]
3787 # links - this is for later, but we need the number of headlines right now
3789 $numMatches = preg_match_all( '/<H(?P<level>[1-6])(?P<attrib>.*?'.'>)(?P<header>.*?)<\/H[1-6] *>/i', $text, $matches );
3791 # if there are fewer than 4 headlines in the article, do not show TOC
3792 # unless it's been explicitly enabled.
3793 $enoughToc = $this->mShowToc
&&
3794 (($numMatches >= 4) ||
$this->mForceTocPosition
);
3796 # Allow user to stipulate that a page should have a "new section"
3797 # link added via __NEWSECTIONLINK__
3798 $mw =& MagicWord
::get( 'newsectionlink' );
3799 if( $mw->matchAndRemove( $text ) )
3800 $this->mOutput
->setNewSection( true );
3802 # if the string __FORCETOC__ (not case-sensitive) occurs in the HTML,
3803 # override above conditions and always show TOC above first header
3804 $mw =& MagicWord
::get( 'forcetoc' );
3805 if ($mw->matchAndRemove( $text ) ) {
3806 $this->mShowToc
= true;
3810 # We need this to perform operations on the HTML
3811 $sk = $this->mOptions
->getSkin();
3817 # Ugh .. the TOC should have neat indentation levels which can be
3818 # passed to the skin functions. These are determined here
3822 $sublevelCount = array();
3823 $levelCount = array();
3829 $markerRegex = "{$this->mUniqPrefix}-h-(\d+)-{$this->mMarkerSuffix}";
3830 $baseTitleText = $this->mTitle
->getPrefixedDBkey();
3833 foreach( $matches[3] as $headline ) {
3834 $isTemplate = false;
3836 $sectionIndex = false;
3838 $markerMatches = array();
3839 if (preg_match("/^$markerRegex/", $headline, $markerMatches)) {
3840 $serial = $markerMatches[1];
3841 list( $titleText, $sectionIndex ) = $this->mHeadings
[$serial];
3842 $isTemplate = ($titleText != $baseTitleText);
3843 $headline = preg_replace("/^$markerRegex/", "", $headline);
3847 $prevlevel = $level;
3848 $prevtoclevel = $toclevel;
3850 $level = $matches[1][$headlineCount];
3852 if( $doNumberHeadings ||
$enoughToc ) {
3854 if ( $level > $prevlevel ) {
3855 # Increase TOC level
3857 $sublevelCount[$toclevel] = 0;
3858 if( $toclevel<$wgMaxTocLevel ) {
3859 $prevtoclevel = $toclevel;
3860 $toc .= $sk->tocIndent();
3864 elseif ( $level < $prevlevel && $toclevel > 1 ) {
3865 # Decrease TOC level, find level to jump to
3867 if ( $toclevel == 2 && $level <= $levelCount[1] ) {
3868 # Can only go down to level 1
3871 for ($i = $toclevel; $i > 0; $i--) {
3872 if ( $levelCount[$i] == $level ) {
3873 # Found last matching level
3877 elseif ( $levelCount[$i] < $level ) {
3878 # Found first matching level below current level
3884 if( $toclevel<$wgMaxTocLevel ) {
3885 if($prevtoclevel < $wgMaxTocLevel) {
3886 # Unindent only if the previous toc level was shown :p
3887 $toc .= $sk->tocUnindent( $prevtoclevel - $toclevel );
3889 $toc .= $sk->tocLineEnd();
3894 # No change in level, end TOC line
3895 if( $toclevel<$wgMaxTocLevel ) {
3896 $toc .= $sk->tocLineEnd();
3900 $levelCount[$toclevel] = $level;
3902 # count number of headlines for each level
3903 @$sublevelCount[$toclevel]++
;
3905 for( $i = 1; $i <= $toclevel; $i++
) {
3906 if( !empty( $sublevelCount[$i] ) ) {
3910 $numbering .= $wgContLang->formatNum( $sublevelCount[$i] );
3916 # The safe header is a version of the header text safe to use for links
3917 # Avoid insertion of weird stuff like <math> by expanding the relevant sections
3918 $safeHeadline = $this->mStripState
->unstripBoth( $headline );
3920 # Remove link placeholders by the link text.
3921 # <!--LINK number-->
3923 # link text with suffix
3924 $safeHeadline = preg_replace( '/<!--LINK ([0-9]*)-->/e',
3925 "\$this->mLinkHolders['texts'][\$1]",
3927 $safeHeadline = preg_replace( '/<!--IWLINK ([0-9]*)-->/e',
3928 "\$this->mInterwikiLinkHolders['texts'][\$1]",
3931 # Strip out HTML (other than plain <sup> and <sub>: bug 8393)
3932 $tocline = preg_replace(
3933 array( '#<(?!/?(sup|sub)).*?'.'>#', '#<(/?(sup|sub)).*?'.'>#' ),
3937 $tocline = trim( $tocline );
3939 # For the anchor, strip out HTML-y stuff period
3940 $safeHeadline = preg_replace( '/<.*?'.'>/', '', $safeHeadline );
3941 $safeHeadline = trim( $safeHeadline );
3943 # Save headline for section edit hint before it's escaped
3944 $headlineHint = $safeHeadline;
3945 $safeHeadline = Sanitizer
::escapeId( $safeHeadline );
3946 $refers[$headlineCount] = $safeHeadline;
3948 # count how many in assoc. array so we can track dupes in anchors
3949 isset( $refers[$safeHeadline] ) ?
$refers[$safeHeadline]++
: $refers[$safeHeadline] = 1;
3950 $refcount[$headlineCount] = $refers[$safeHeadline];
3952 # Don't number the heading if it is the only one (looks silly)
3953 if( $doNumberHeadings && count( $matches[3] ) > 1) {
3954 # the two are different if the line contains a link
3955 $headline=$numbering . ' ' . $headline;
3958 # Create the anchor for linking from the TOC to the section
3959 $anchor = $safeHeadline;
3960 if($refcount[$headlineCount] > 1 ) {
3961 $anchor .= '_' . $refcount[$headlineCount];
3963 if( $enoughToc && ( !isset($wgMaxTocLevel) ||
$toclevel<$wgMaxTocLevel ) ) {
3964 $toc .= $sk->tocLine($anchor, $tocline, $numbering, $toclevel);
3965 $tocraw[] = array( 'toclevel' => $toclevel, 'level' => $level, 'line' => $tocline, 'number' => $numbering );
3967 # give headline the correct <h#> tag
3968 if( $showEditLink && $sectionIndex !== false ) {
3970 # Put a T flag in the section identifier, to indicate to extractSections()
3971 # that sections inside <includeonly> should be counted.
3972 $editlink = $sk->editSectionLinkForOther($titleText, "T-$sectionIndex");
3974 $editlink = $sk->editSectionLink($this->mTitle
, $sectionIndex, $headlineHint);
3979 $head[$headlineCount] = $sk->makeHeadline( $level, $matches['attrib'][$headlineCount], $anchor, $headline, $editlink );
3984 $this->mOutput
->setSections( $tocraw );
3986 # Never ever show TOC if no headers
3987 if( $numVisible < 1 ) {
3992 if( $prevtoclevel > 0 && $prevtoclevel < $wgMaxTocLevel ) {
3993 $toc .= $sk->tocUnindent( $prevtoclevel - 1 );
3995 $toc = $sk->tocList( $toc );
3998 # split up and insert constructed headlines
4000 $blocks = preg_split( '/<H[1-6].*?' . '>.*?<\/H[1-6]>/i', $text );
4003 foreach( $blocks as $block ) {
4004 if( $showEditLink && $headlineCount > 0 && $i == 0 && $block != "\n" ) {
4005 # This is the [edit] link that appears for the top block of text when
4006 # section editing is enabled
4008 # Disabled because it broke block formatting
4009 # For example, a bullet point in the top line
4010 # $full .= $sk->editSectionLink(0);
4013 if( $enoughToc && !$i && $isMain && !$this->mForceTocPosition
) {
4014 # Top anchor now in skin
4018 if( !empty( $head[$i] ) ) {
4023 if( $this->mForceTocPosition
) {
4024 return str_replace( '<!--MWTOC-->', $toc, $full );
4031 * Transform wiki markup when saving a page by doing \r\n -> \n
4032 * conversion, substitting signatures, {{subst:}} templates, etc.
4034 * @param string $text the text to transform
4035 * @param Title &$title the Title object for the current article
4036 * @param User &$user the User object describing the current user
4037 * @param ParserOptions $options parsing options
4038 * @param bool $clearState whether to clear the parser state first
4039 * @return string the altered wiki markup
4042 function preSaveTransform( $text, &$title, $user, $options, $clearState = true ) {
4043 $this->mOptions
= $options;
4044 $this->mTitle
=& $title;
4045 $this->setOutputType( OT_WIKI
);
4047 if ( $clearState ) {
4048 $this->clearState();
4054 $text = str_replace( array_keys( $pairs ), array_values( $pairs ), $text );
4055 $text = $this->pstPass2( $text, $user );
4056 $text = $this->mStripState
->unstripBoth( $text );
4061 * Pre-save transform helper function
4064 function pstPass2( $text, $user ) {
4065 global $wgContLang, $wgLocaltimezone;
4067 /* Note: This is the timestamp saved as hardcoded wikitext to
4068 * the database, we use $wgContLang here in order to give
4069 * everyone the same signature and use the default one rather
4070 * than the one selected in each user's preferences.
4072 if ( isset( $wgLocaltimezone ) ) {
4073 $oldtz = getenv( 'TZ' );
4074 putenv( 'TZ='.$wgLocaltimezone );
4076 $d = $wgContLang->timeanddate( date( 'YmdHis' ), false, false) .
4077 ' (' . date( 'T' ) . ')';
4078 if ( isset( $wgLocaltimezone ) ) {
4079 putenv( 'TZ='.$oldtz );
4082 # Variable replacement
4083 # Because mOutputType is OT_WIKI, this will only process {{subst:xxx}} type tags
4084 $text = $this->replaceVariables( $text );
4086 # Strip out <nowiki> etc. added via replaceVariables
4087 #$text = $this->strip( $text, $this->mStripState, false, array( 'gallery' ) );
4090 $sigText = $this->getUserSig( $user );
4091 $text = strtr( $text, array(
4093 '~~~~' => "$sigText $d",
4097 # Context links: [[|name]] and [[name (context)|]]
4099 global $wgLegalTitleChars;
4100 $tc = "[$wgLegalTitleChars]";
4101 $nc = '[ _0-9A-Za-z\x80-\xff]'; # Namespaces can use non-ascii!
4103 $p1 = "/\[\[(:?$nc+:|:|)($tc+?)( \\($tc+\\))\\|]]/"; # [[ns:page (context)|]]
4104 $p3 = "/\[\[(:?$nc+:|:|)($tc+?)( \\($tc+\\)|)(, $tc+|)\\|]]/"; # [[ns:page (context), context|]]
4105 $p2 = "/\[\[\\|($tc+)]]/"; # [[|page]]
4107 # try $p1 first, to turn "[[A, B (C)|]]" into "[[A, B (C)|A, B]]"
4108 $text = preg_replace( $p1, '[[\\1\\2\\3|\\2]]', $text );
4109 $text = preg_replace( $p3, '[[\\1\\2\\3\\4|\\2]]', $text );
4111 $t = $this->mTitle
->getText();
4113 if ( preg_match( "/^($nc+:|)$tc+?( \\($tc+\\))$/", $t, $m ) ) {
4114 $text = preg_replace( $p2, "[[$m[1]\\1$m[2]|\\1]]", $text );
4115 } elseif ( preg_match( "/^($nc+:|)$tc+?(, $tc+|)$/", $t, $m ) && '' != "$m[1]$m[2]" ) {
4116 $text = preg_replace( $p2, "[[$m[1]\\1$m[2]|\\1]]", $text );
4118 # if there's no context, don't bother duplicating the title
4119 $text = preg_replace( $p2, '[[\\1]]', $text );
4122 # Trim trailing whitespace
4123 $text = rtrim( $text );
4129 * Fetch the user's signature text, if any, and normalize to
4130 * validated, ready-to-insert wikitext.
4136 function getUserSig( &$user ) {
4137 global $wgMaxSigChars;
4139 $username = $user->getName();
4140 $nickname = $user->getOption( 'nickname' );
4141 $nickname = $nickname === '' ?
$username : $nickname;
4143 if( mb_strlen( $nickname ) > $wgMaxSigChars ) {
4144 $nickname = $username;
4145 wfDebug( __METHOD__
. ": $username has overlong signature.\n" );
4146 } elseif( $user->getBoolOption( 'fancysig' ) !== false ) {
4147 # Sig. might contain markup; validate this
4148 if( $this->validateSig( $nickname ) !== false ) {
4149 # Validated; clean up (if needed) and return it
4150 return $this->cleanSig( $nickname, true );
4152 # Failed to validate; fall back to the default
4153 $nickname = $username;
4154 wfDebug( "Parser::getUserSig: $username has bad XML tags in signature.\n" );
4158 // Make sure nickname doesnt get a sig in a sig
4159 $nickname = $this->cleanSigInSig( $nickname );
4161 # If we're still here, make it a link to the user page
4162 $userText = wfEscapeWikiText( $username );
4163 $nickText = wfEscapeWikiText( $nickname );
4164 if ( $user->isAnon() ) {
4165 return wfMsgExt( 'signature-anon', array( 'content', 'parsemag' ), $userText, $nickText );
4167 return wfMsgExt( 'signature', array( 'content', 'parsemag' ), $userText, $nickText );
4172 * Check that the user's signature contains no bad XML
4174 * @param string $text
4175 * @return mixed An expanded string, or false if invalid.
4177 function validateSig( $text ) {
4178 return( wfIsWellFormedXmlFragment( $text ) ?
$text : false );
4182 * Clean up signature text
4184 * 1) Strip ~~~, ~~~~ and ~~~~~ out of signatures @see cleanSigInSig
4185 * 2) Substitute all transclusions
4187 * @param string $text
4188 * @param $parsing Whether we're cleaning (preferences save) or parsing
4189 * @return string Signature text
4191 function cleanSig( $text, $parsing = false ) {
4194 $this->startExternalParse( $wgTitle, new ParserOptions(), OT_MSG
);
4197 # FIXME: regex doesn't respect extension tags or nowiki
4198 # => Move this logic to braceSubstitution()
4199 $substWord = MagicWord
::get( 'subst' );
4200 $substRegex = '/\{\{(?!(?:' . $substWord->getBaseRegex() . '))/x' . $substWord->getRegexCase();
4201 $substText = '{{' . $substWord->getSynonym( 0 );
4203 $text = preg_replace( $substRegex, $substText, $text );
4204 $text = $this->cleanSigInSig( $text );
4205 $dom = $this->preprocessToDom( $text );
4206 $frame = new PPFrame( $this );
4207 $text = $frame->expand( $dom->documentElement
);
4210 $text = $this->mStripState
->unstripBoth( $text );
4217 * Strip ~~~, ~~~~ and ~~~~~ out of signatures
4218 * @param string $text
4219 * @return string Signature text with /~{3,5}/ removed
4221 function cleanSigInSig( $text ) {
4222 $text = preg_replace( '/~{3,5}/', '', $text );
4227 * Set up some variables which are usually set up in parse()
4228 * so that an external function can call some class members with confidence
4231 function startExternalParse( &$title, $options, $outputType, $clearState = true ) {
4232 $this->mTitle
=& $title;
4233 $this->mOptions
= $options;
4234 $this->setOutputType( $outputType );
4235 if ( $clearState ) {
4236 $this->clearState();
4241 * Transform a MediaWiki message by replacing magic variables.
4243 * For some unknown reason, it also expands templates, but only to the
4244 * first recursion level. This is wrong and broken, probably introduced
4245 * accidentally during refactoring, but probably relied upon by thousands
4248 * @param string $text the text to transform
4249 * @param ParserOptions $options options
4250 * @return string the text with variables substituted
4253 function transformMsg( $text, $options ) {
4255 static $executing = false;
4257 $fname = "Parser::transformMsg";
4259 # Guard against infinite recursion
4265 wfProfileIn($fname);
4267 if ( $wgTitle && !( $wgTitle instanceof FakeTitle
) ) {
4268 $this->mTitle
= $wgTitle;
4270 $this->mTitle
= Title
::newFromText('msg');
4272 $this->mOptions
= $options;
4273 $this->setOutputType( OT_MSG
);
4274 $this->clearState();
4275 $text = $this->replaceVariables( $text );
4276 $text = $this->mStripState
->unstripBoth( $text );
4279 wfProfileOut($fname);
4284 * Create an HTML-style tag, e.g. <yourtag>special text</yourtag>
4285 * The callback should have the following form:
4286 * function myParserHook( $text, $params, &$parser ) { ... }
4288 * Transform and return $text. Use $parser for any required context, e.g. use
4289 * $parser->getTitle() and $parser->getOptions() not $wgTitle or $wgOut->mParserOptions
4293 * @param mixed $tag The tag to use, e.g. 'hook' for <hook>
4294 * @param mixed $callback The callback function (and object) to use for the tag
4296 * @return The old value of the mTagHooks array associated with the hook
4298 function setHook( $tag, $callback ) {
4299 $tag = strtolower( $tag );
4300 $oldVal = isset( $this->mTagHooks
[$tag] ) ?
$this->mTagHooks
[$tag] : null;
4301 $this->mTagHooks
[$tag] = $callback;
4302 $this->mStripList
[] = $tag;
4307 function setTransparentTagHook( $tag, $callback ) {
4308 $tag = strtolower( $tag );
4309 $oldVal = isset( $this->mTransparentTagHooks
[$tag] ) ?
$this->mTransparentTagHooks
[$tag] : null;
4310 $this->mTransparentTagHooks
[$tag] = $callback;
4316 * Create a function, e.g. {{sum:1|2|3}}
4317 * The callback function should have the form:
4318 * function myParserFunction( &$parser, $arg1, $arg2, $arg3 ) { ... }
4320 * The callback may either return the text result of the function, or an array with the text
4321 * in element 0, and a number of flags in the other elements. The names of the flags are
4322 * specified in the keys. Valid flags are:
4323 * found The text returned is valid, stop processing the template. This
4325 * nowiki Wiki markup in the return value should be escaped
4326 * noparse Unsafe HTML tags should not be stripped, etc.
4327 * noargs Don't replace triple-brace arguments in the return value
4328 * isHTML The returned text is HTML, armour it against wikitext transformation
4332 * @param string $id The magic word ID
4333 * @param mixed $callback The callback function (and object) to use
4334 * @param integer $flags a combination of the following flags:
4335 * SFH_NO_HASH No leading hash, i.e. {{plural:...}} instead of {{#if:...}}
4337 * @return The old callback function for this name, if any
4339 function setFunctionHook( $id, $callback, $flags = 0 ) {
4340 $oldVal = isset( $this->mFunctionHooks
[$id] ) ?
$this->mFunctionHooks
[$id][0] : null;
4341 $this->mFunctionHooks
[$id] = array( $callback, $flags );
4343 # Add to function cache
4344 $mw = MagicWord
::get( $id );
4346 throw new MWException( 'Parser::setFunctionHook() expecting a magic word identifier.' );
4348 $synonyms = $mw->getSynonyms();
4349 $sensitive = intval( $mw->isCaseSensitive() );
4351 foreach ( $synonyms as $syn ) {
4353 if ( !$sensitive ) {
4354 $syn = strtolower( $syn );
4357 if ( !( $flags & SFH_NO_HASH
) ) {
4360 # Remove trailing colon
4361 if ( substr( $syn, -1, 1 ) == ':' ) {
4362 $syn = substr( $syn, 0, -1 );
4364 $this->mFunctionSynonyms
[$sensitive][$syn] = $id;
4370 * Get all registered function hook identifiers
4374 function getFunctionHooks() {
4375 return array_keys( $this->mFunctionHooks
);
4379 * Replace <!--LINK--> link placeholders with actual links, in the buffer
4380 * Placeholders created in Skin::makeLinkObj()
4381 * Returns an array of link CSS classes, indexed by PDBK.
4382 * $options is a bit field, RLH_FOR_UPDATE to select for update
4384 function replaceLinkHolders( &$text, $options = 0 ) {
4388 $fname = 'Parser::replaceLinkHolders';
4389 wfProfileIn( $fname );
4393 $linkcolour_ids = array();
4394 $sk = $this->mOptions
->getSkin();
4395 $linkCache =& LinkCache
::singleton();
4397 if ( !empty( $this->mLinkHolders
['namespaces'] ) ) {
4398 wfProfileIn( $fname.'-check' );
4399 $dbr = wfGetDB( DB_SLAVE
);
4400 $page = $dbr->tableName( 'page' );
4401 $threshold = $wgUser->getOption('stubthreshold');
4404 asort( $this->mLinkHolders
['namespaces'] );
4409 foreach ( $this->mLinkHolders
['namespaces'] as $key => $ns ) {
4411 $title = $this->mLinkHolders
['titles'][$key];
4413 # Skip invalid entries.
4414 # Result will be ugly, but prevents crash.
4415 if ( is_null( $title ) ) {
4418 $pdbk = $pdbks[$key] = $title->getPrefixedDBkey();
4420 # Check if it's a static known link, e.g. interwiki
4421 if ( $title->isAlwaysKnown() ) {
4422 $colours[$pdbk] = '';
4423 } elseif ( ( $id = $linkCache->getGoodLinkID( $pdbk ) ) != 0 ) {
4424 $colours[$pdbk] = '';
4425 $this->mOutput
->addLink( $title, $id );
4426 } elseif ( $linkCache->isBadLink( $pdbk ) ) {
4427 $colours[$pdbk] = 'new';
4428 } elseif ( $title->getNamespace() == NS_SPECIAL
&& !SpecialPage
::exists( $pdbk ) ) {
4429 $colours[$pdbk] = 'new';
4431 # Not in the link cache, add it to the query
4432 if ( !isset( $current ) ) {
4434 $query = "SELECT page_id, page_namespace, page_title";
4435 if ( $threshold > 0 ) {
4436 $query .= ', page_len, page_is_redirect';
4438 $query .= " FROM $page WHERE (page_namespace=$ns AND page_title IN(";
4439 } elseif ( $current != $ns ) {
4441 $query .= ")) OR (page_namespace=$ns AND page_title IN(";
4446 $query .= $dbr->addQuotes( $this->mLinkHolders
['dbkeys'][$key] );
4451 if ( $options & RLH_FOR_UPDATE
) {
4452 $query .= ' FOR UPDATE';
4455 $res = $dbr->query( $query, $fname );
4457 # Fetch data and form into an associative array
4458 # non-existent = broken
4459 while ( $s = $dbr->fetchObject($res) ) {
4460 $title = Title
::makeTitle( $s->page_namespace
, $s->page_title
);
4461 $pdbk = $title->getPrefixedDBkey();
4462 $linkCache->addGoodLinkObj( $s->page_id
, $title );
4463 $this->mOutput
->addLink( $title, $s->page_id
);
4464 $colours[$pdbk] = $sk->getLinkColour( $s, $threshold );
4465 //add id to the extension todolist
4466 $linkcolour_ids[$s->page_id
] = $pdbk;
4468 //pass an array of page_ids to an extension
4469 wfRunHooks( 'GetLinkColours', array( $linkcolour_ids, &$colours ) );
4471 wfProfileOut( $fname.'-check' );
4473 # Do a second query for different language variants of links and categories
4474 if($wgContLang->hasVariants()){
4475 $linkBatch = new LinkBatch();
4476 $variantMap = array(); // maps $pdbkey_Variant => $keys (of link holders)
4477 $categoryMap = array(); // maps $category_variant => $category (dbkeys)
4478 $varCategories = array(); // category replacements oldDBkey => newDBkey
4480 $categories = $this->mOutput
->getCategoryLinks();
4482 // Add variants of links to link batch
4483 foreach ( $this->mLinkHolders
['namespaces'] as $key => $ns ) {
4484 $title = $this->mLinkHolders
['titles'][$key];
4485 if ( is_null( $title ) )
4488 $pdbk = $title->getPrefixedDBkey();
4489 $titleText = $title->getText();
4491 // generate all variants of the link title text
4492 $allTextVariants = $wgContLang->convertLinkToAllVariants($titleText);
4494 // if link was not found (in first query), add all variants to query
4495 if ( !isset($colours[$pdbk]) ){
4496 foreach($allTextVariants as $textVariant){
4497 if($textVariant != $titleText){
4498 $variantTitle = Title
::makeTitle( $ns, $textVariant );
4499 if(is_null($variantTitle)) continue;
4500 $linkBatch->addObj( $variantTitle );
4501 $variantMap[$variantTitle->getPrefixedDBkey()][] = $key;
4507 // process categories, check if a category exists in some variant
4508 foreach( $categories as $category ){
4509 $variants = $wgContLang->convertLinkToAllVariants($category);
4510 foreach($variants as $variant){
4511 if($variant != $category){
4512 $variantTitle = Title
::newFromDBkey( Title
::makeName(NS_CATEGORY
,$variant) );
4513 if(is_null($variantTitle)) continue;
4514 $linkBatch->addObj( $variantTitle );
4515 $categoryMap[$variant] = $category;
4521 if(!$linkBatch->isEmpty()){
4523 $titleClause = $linkBatch->constructSet('page', $dbr);
4525 $variantQuery = "SELECT page_id, page_namespace, page_title";
4526 if ( $threshold > 0 ) {
4527 $variantQuery .= ', page_len, page_is_redirect';
4530 $variantQuery .= " FROM $page WHERE $titleClause";
4531 if ( $options & RLH_FOR_UPDATE
) {
4532 $variantQuery .= ' FOR UPDATE';
4535 $varRes = $dbr->query( $variantQuery, $fname );
4537 // for each found variants, figure out link holders and replace
4538 while ( $s = $dbr->fetchObject($varRes) ) {
4540 $variantTitle = Title
::makeTitle( $s->page_namespace
, $s->page_title
);
4541 $varPdbk = $variantTitle->getPrefixedDBkey();
4542 $vardbk = $variantTitle->getDBkey();
4544 $holderKeys = array();
4545 if(isset($variantMap[$varPdbk])){
4546 $holderKeys = $variantMap[$varPdbk];
4547 $linkCache->addGoodLinkObj( $s->page_id
, $variantTitle );
4548 $this->mOutput
->addLink( $variantTitle, $s->page_id
);
4551 // loop over link holders
4552 foreach($holderKeys as $key){
4553 $title = $this->mLinkHolders
['titles'][$key];
4554 if ( is_null( $title ) ) continue;
4556 $pdbk = $title->getPrefixedDBkey();
4558 if(!isset($colours[$pdbk])){
4559 // found link in some of the variants, replace the link holder data
4560 $this->mLinkHolders
['titles'][$key] = $variantTitle;
4561 $this->mLinkHolders
['dbkeys'][$key] = $variantTitle->getDBkey();
4563 // set pdbk and colour
4564 $pdbks[$key] = $varPdbk;
4565 $colours[$varPdbk] = $sk->getLinkColour( $s, $threshold );
4566 $linkcolour_ids[$s->page_id
] = $pdbk;
4568 wfRunHooks( 'GetLinkColours', array( $linkcolour_ids, &$colours ) );
4571 // check if the object is a variant of a category
4572 if(isset($categoryMap[$vardbk])){
4573 $oldkey = $categoryMap[$vardbk];
4574 if($oldkey != $vardbk)
4575 $varCategories[$oldkey]=$vardbk;
4579 // rebuild the categories in original order (if there are replacements)
4580 if(count($varCategories)>0){
4582 $originalCats = $this->mOutput
->getCategories();
4583 foreach($originalCats as $cat => $sortkey){
4584 // make the replacement
4585 if( array_key_exists($cat,$varCategories) )
4586 $newCats[$varCategories[$cat]] = $sortkey;
4587 else $newCats[$cat] = $sortkey;
4589 $this->mOutput
->setCategoryLinks($newCats);
4594 # Construct search and replace arrays
4595 wfProfileIn( $fname.'-construct' );
4596 $replacePairs = array();
4597 foreach ( $this->mLinkHolders
['namespaces'] as $key => $ns ) {
4598 $pdbk = $pdbks[$key];
4599 $searchkey = "<!--LINK $key-->";
4600 $title = $this->mLinkHolders
['titles'][$key];
4601 if ( !isset( $colours[$pdbk] ) ||
$colours[$pdbk] == 'new' ) {
4602 $linkCache->addBadLinkObj( $title );
4603 $colours[$pdbk] = 'new';
4604 $this->mOutput
->addLink( $title, 0 );
4605 $replacePairs[$searchkey] = $sk->makeBrokenLinkObj( $title,
4606 $this->mLinkHolders
['texts'][$key],
4607 $this->mLinkHolders
['queries'][$key] );
4609 $replacePairs[$searchkey] = $sk->makeColouredLinkObj( $title, $colours[$pdbk],
4610 $this->mLinkHolders
['texts'][$key],
4611 $this->mLinkHolders
['queries'][$key] );
4614 $replacer = new HashtableReplacer( $replacePairs, 1 );
4615 wfProfileOut( $fname.'-construct' );
4618 wfProfileIn( $fname.'-replace' );
4619 $text = preg_replace_callback(
4620 '/(<!--LINK .*?-->)/',
4624 wfProfileOut( $fname.'-replace' );
4627 # Now process interwiki link holders
4628 # This is quite a bit simpler than internal links
4629 if ( !empty( $this->mInterwikiLinkHolders
['texts'] ) ) {
4630 wfProfileIn( $fname.'-interwiki' );
4631 # Make interwiki link HTML
4632 $replacePairs = array();
4633 foreach( $this->mInterwikiLinkHolders
['texts'] as $key => $link ) {
4634 $title = $this->mInterwikiLinkHolders
['titles'][$key];
4635 $replacePairs[$key] = $sk->makeLinkObj( $title, $link );
4637 $replacer = new HashtableReplacer( $replacePairs, 1 );
4639 $text = preg_replace_callback(
4640 '/<!--IWLINK (.*?)-->/',
4643 wfProfileOut( $fname.'-interwiki' );
4646 wfProfileOut( $fname );
4651 * Replace <!--LINK--> link placeholders with plain text of links
4652 * (not HTML-formatted).
4653 * @param string $text
4656 function replaceLinkHoldersText( $text ) {
4657 $fname = 'Parser::replaceLinkHoldersText';
4658 wfProfileIn( $fname );
4660 $text = preg_replace_callback(
4661 '/<!--(LINK|IWLINK) (.*?)-->/',
4662 array( &$this, 'replaceLinkHoldersTextCallback' ),
4665 wfProfileOut( $fname );
4670 * @param array $matches
4674 function replaceLinkHoldersTextCallback( $matches ) {
4675 $type = $matches[1];
4677 if( $type == 'LINK' ) {
4678 if( isset( $this->mLinkHolders
['texts'][$key] ) ) {
4679 return $this->mLinkHolders
['texts'][$key];
4681 } elseif( $type == 'IWLINK' ) {
4682 if( isset( $this->mInterwikiLinkHolders
['texts'][$key] ) ) {
4683 return $this->mInterwikiLinkHolders
['texts'][$key];
4690 * Tag hook handler for 'pre'.
4692 function renderPreTag( $text, $attribs ) {
4693 // Backwards-compatibility hack
4694 $content = StringUtils
::delimiterReplace( '<nowiki>', '</nowiki>', '$1', $text, 'i' );
4696 $attribs = Sanitizer
::validateTagAttributes( $attribs, 'pre' );
4697 return wfOpenElement( 'pre', $attribs ) .
4698 Xml
::escapeTagsOnly( $content ) .
4703 * Renders an image gallery from a text with one line per image.
4704 * text labels may be given by using |-style alternative text. E.g.
4705 * Image:one.jpg|The number "1"
4706 * Image:tree.jpg|A tree
4707 * given as text will return the HTML of a gallery with two images,
4708 * labeled 'The number "1"' and
4711 function renderImageGallery( $text, $params ) {
4712 $ig = new ImageGallery();
4713 $ig->setContextTitle( $this->mTitle
);
4714 $ig->setShowBytes( false );
4715 $ig->setShowFilename( false );
4716 $ig->setParser( $this );
4717 $ig->setHideBadImages();
4718 $ig->setAttributes( Sanitizer
::validateTagAttributes( $params, 'table' ) );
4719 $ig->useSkin( $this->mOptions
->getSkin() );
4720 $ig->mRevisionId
= $this->mRevisionId
;
4722 if( isset( $params['caption'] ) ) {
4723 $caption = $params['caption'];
4724 $caption = htmlspecialchars( $caption );
4725 $caption = $this->replaceInternalLinks( $caption );
4726 $ig->setCaptionHtml( $caption );
4728 if( isset( $params['perrow'] ) ) {
4729 $ig->setPerRow( $params['perrow'] );
4731 if( isset( $params['widths'] ) ) {
4732 $ig->setWidths( $params['widths'] );
4734 if( isset( $params['heights'] ) ) {
4735 $ig->setHeights( $params['heights'] );
4738 wfRunHooks( 'BeforeParserrenderImageGallery', array( &$this, &$ig ) );
4740 $lines = explode( "\n", $text );
4741 foreach ( $lines as $line ) {
4742 # match lines like these:
4743 # Image:someimage.jpg|This is some image
4745 preg_match( "/^([^|]+)(\\|(.*))?$/", $line, $matches );
4747 if ( count( $matches ) == 0 ) {
4750 $tp = Title
::newFromText( $matches[1] );
4752 if( is_null( $nt ) ) {
4753 # Bogus title. Ignore these so we don't bomb out later.
4756 if ( isset( $matches[3] ) ) {
4757 $label = $matches[3];
4762 $pout = $this->parse( $label,
4765 false, // Strip whitespace...?
4766 false // Don't clear state!
4768 $html = $pout->getText();
4770 $ig->add( $nt, $html );
4772 # Only add real images (bug #5586)
4773 if ( $nt->getNamespace() == NS_IMAGE
) {
4774 $this->mOutput
->addImage( $nt->getDBkey() );
4777 return $ig->toHTML();
4780 function getImageParams( $handler ) {
4782 $handlerClass = get_class( $handler );
4786 if ( !isset( $this->mImageParams
[$handlerClass] ) ) {
4787 // Initialise static lists
4788 static $internalParamNames = array(
4789 'horizAlign' => array( 'left', 'right', 'center', 'none' ),
4790 'vertAlign' => array( 'baseline', 'sub', 'super', 'top', 'text-top', 'middle',
4791 'bottom', 'text-bottom' ),
4792 'frame' => array( 'thumbnail', 'manualthumb', 'framed', 'frameless',
4793 'upright', 'border' ),
4795 static $internalParamMap;
4796 if ( !$internalParamMap ) {
4797 $internalParamMap = array();
4798 foreach ( $internalParamNames as $type => $names ) {
4799 foreach ( $names as $name ) {
4800 $magicName = str_replace( '-', '_', "img_$name" );
4801 $internalParamMap[$magicName] = array( $type, $name );
4806 // Add handler params
4807 $paramMap = $internalParamMap;
4809 $handlerParamMap = $handler->getParamMap();
4810 foreach ( $handlerParamMap as $magic => $paramName ) {
4811 $paramMap[$magic] = array( 'handler', $paramName );
4814 $this->mImageParams
[$handlerClass] = $paramMap;
4815 $this->mImageParamsMagicArray
[$handlerClass] = new MagicWordArray( array_keys( $paramMap ) );
4817 return array( $this->mImageParams
[$handlerClass], $this->mImageParamsMagicArray
[$handlerClass] );
4821 * Parse image options text and use it to make an image
4823 function makeImage( $title, $options ) {
4824 # @TODO: let the MediaHandler specify its transform parameters
4826 # Check if the options text is of the form "options|alt text"
4828 # * thumbnail make a thumbnail with enlarge-icon and caption, alignment depends on lang
4829 # * left no resizing, just left align. label is used for alt= only
4830 # * right same, but right aligned
4831 # * none same, but not aligned
4832 # * ___px scale to ___ pixels width, no aligning. e.g. use in taxobox
4833 # * center center the image
4834 # * framed Keep original image size, no magnify-button.
4835 # * frameless like 'thumb' but without a frame. Keeps user preferences for width
4836 # * upright reduce width for upright images, rounded to full __0 px
4837 # * border draw a 1px border around the image
4838 # vertical-align values (no % or length right now):
4848 $parts = array_map( 'trim', explode( '|', $options) );
4849 $sk = $this->mOptions
->getSkin();
4851 # Give extensions a chance to select the file revision for us
4852 $skip = $time = false;
4853 wfRunHooks( 'BeforeParserMakeImageLinkObj', array( &$this, &$title, &$skip, &$time ) );
4856 return $sk->makeLinkObj( $title );
4860 $file = wfFindFile( $title, $time );
4861 $handler = $file ?
$file->getHandler() : false;
4863 list( $paramMap, $mwArray ) = $this->getImageParams( $handler );
4865 # Process the input parameters
4867 $params = array( 'frame' => array(), 'handler' => array(),
4868 'horizAlign' => array(), 'vertAlign' => array() );
4869 foreach( $parts as $part ) {
4870 list( $magicName, $value ) = $mwArray->matchVariableStartToEnd( $part );
4871 if ( isset( $paramMap[$magicName] ) ) {
4872 list( $type, $paramName ) = $paramMap[$magicName];
4873 $params[$type][$paramName] = $value;
4875 // Special case; width and height come in one variable together
4876 if( $type == 'handler' && $paramName == 'width' ) {
4878 if ( preg_match( '/^([0-9]*)x([0-9]*)$/', $value, $m ) ) {
4879 $params[$type]['width'] = intval( $m[1] );
4880 $params[$type]['height'] = intval( $m[2] );
4882 $params[$type]['width'] = intval( $value );
4890 # Process alignment parameters
4891 if ( $params['horizAlign'] ) {
4892 $params['frame']['align'] = key( $params['horizAlign'] );
4894 if ( $params['vertAlign'] ) {
4895 $params['frame']['valign'] = key( $params['vertAlign'] );
4898 # Validate the handler parameters
4900 foreach ( $params['handler'] as $name => $value ) {
4901 if ( !$handler->validateParam( $name, $value ) ) {
4902 unset( $params['handler'][$name] );
4907 # Strip bad stuff out of the alt text
4908 $alt = $this->replaceLinkHoldersText( $caption );
4910 # make sure there are no placeholders in thumbnail attributes
4911 # that are later expanded to html- so expand them now and
4913 $alt = $this->mStripState
->unstripBoth( $alt );
4914 $alt = Sanitizer
::stripAllTags( $alt );
4916 $params['frame']['alt'] = $alt;
4917 $params['frame']['caption'] = $caption;
4919 # Linker does the rest
4920 $ret = $sk->makeImageLink2( $title, $file, $params['frame'], $params['handler'] );
4922 # Give the handler a chance to modify the parser object
4924 $handler->parserTransformHook( $this, $file );
4931 * Set a flag in the output object indicating that the content is dynamic and
4932 * shouldn't be cached.
4934 function disableCache() {
4935 wfDebug( "Parser output marked as uncacheable.\n" );
4936 $this->mOutput
->mCacheTime
= -1;
4940 * Callback from the Sanitizer for expanding items found in HTML attribute
4941 * values, so they can be safely tested and escaped.
4942 * @param string $text
4943 * @param PPFrame $frame
4947 function attributeStripCallback( &$text, $frame = false ) {
4948 $text = $this->replaceVariables( $text, $frame );
4949 $text = $this->mStripState
->unstripBoth( $text );
4958 function Title( $x = NULL ) { return wfSetVar( $this->mTitle
, $x ); }
4959 function Options( $x = NULL ) { return wfSetVar( $this->mOptions
, $x ); }
4960 function OutputType( $x = NULL ) { return wfSetVar( $this->mOutputType
, $x ); }
4966 function getTags() { return array_merge( array_keys($this->mTransparentTagHooks
), array_keys( $this->mTagHooks
) ); }
4971 * Break wikitext input into sections, and either pull or replace
4972 * some particular section's text.
4974 * External callers should use the getSection and replaceSection methods.
4976 * @param string $text Page wikitext
4977 * @param string $section A section identifier string of the form:
4978 * <flag1> - <flag2> - ... - <section number>
4980 * Currently the only recognised flag is "T", which means the target section number
4981 * was derived during a template inclusion parse, in other words this is a template
4982 * section edit link. If no flags are given, it was an ordinary section edit link.
4983 * This flag is required to avoid a section numbering mismatch when a section is
4984 * enclosed by <includeonly> (bug 6563).
4986 * The section number 0 pulls the text before the first heading; other numbers will
4987 * pull the given section along with its lower-level subsections. If the section is
4988 * not found, $mode=get will return $newtext, and $mode=replace will return $text.
4990 * @param string $mode One of "get" or "replace"
4991 * @param string $newText Replacement text for section data.
4992 * @return string for "get", the extracted section text.
4993 * for "replace", the whole page with the section replaced.
4995 private function extractSections( $text, $section, $mode, $newText='' ) {
4997 $this->clearState();
4998 $this->mTitle
= $wgTitle; // not generally used but removes an ugly failure mode
4999 $this->mOptions
= new ParserOptions
;
5000 $this->setOutputType( OT_WIKI
);
5003 $frame = new PPFrame( $this );
5005 // Process section extraction flags
5007 $sectionParts = explode( '-', $section );
5008 $sectionIndex = array_pop( $sectionParts );
5009 foreach ( $sectionParts as $part ) {
5010 if ( $part == 'T' ) {
5011 $flags |
= self
::PTD_FOR_INCLUSION
;
5014 // Preprocess the text
5015 $dom = $this->preprocessToDom( $text, $flags );
5016 $root = $dom->documentElement
;
5018 // <h> nodes indicate section breaks
5019 // They can only occur at the top level, so we can find them by iterating the root's children
5020 $node = $root->firstChild
;
5022 // Find the target section
5023 if ( $sectionIndex == 0 ) {
5024 // Section zero doesn't nest, level=big
5025 $targetLevel = 1000;
5028 if ( $node->nodeName
== 'h' ) {
5029 if ( $curIndex +
1 == $sectionIndex ) {
5034 if ( $mode == 'replace' ) {
5035 $outText .= $frame->expand( $node, PPFrame
::RECOVER_ORIG
);
5037 $node = $node->nextSibling
;
5040 $targetLevel = $node->getAttribute( 'level' );
5046 if ( $mode == 'get' ) {
5053 // Find the end of the section, including nested sections
5055 if ( $node->nodeName
== 'h' ) {
5057 $curLevel = $node->getAttribute( 'level' );
5058 if ( $curIndex != $sectionIndex && $curLevel <= $targetLevel ) {
5062 if ( $mode == 'get' ) {
5063 $outText .= $frame->expand( $node, PPFrame
::RECOVER_ORIG
);
5065 $node = $node->nextSibling
;
5068 // Write out the remainder (in replace mode only)
5069 if ( $mode == 'replace' ) {
5070 // Output the replacement text
5071 // Add two newlines on -- trailing whitespace in $newText is conventionally
5072 // stripped by the editor, so we need both newlines to restore the paragraph gap
5073 $outText .= $newText . "\n\n";
5075 $outText .= $frame->expand( $node, PPFrame
::RECOVER_ORIG
);
5076 $node = $node->nextSibling
;
5080 if ( is_string( $outText ) ) {
5081 // Re-insert stripped tags
5082 $outText = trim( $this->mStripState
->unstripBoth( $outText ) );
5089 * This function returns the text of a section, specified by a number ($section).
5090 * A section is text under a heading like == Heading == or \<h1\>Heading\</h1\>, or
5091 * the first section before any such heading (section 0).
5093 * If a section contains subsections, these are also returned.
5095 * @param string $text text to look in
5096 * @param string $section section identifier
5097 * @param string $deftext default to return if section is not found
5098 * @return string text of the requested section
5100 public function getSection( $text, $section, $deftext='' ) {
5101 return $this->extractSections( $text, $section, "get", $deftext );
5104 public function replaceSection( $oldtext, $section, $text ) {
5105 return $this->extractSections( $oldtext, $section, "replace", $text );
5109 * Get the timestamp associated with the current revision, adjusted for
5110 * the default server-local timestamp
5112 function getRevisionTimestamp() {
5113 if ( is_null( $this->mRevisionTimestamp
) ) {
5114 wfProfileIn( __METHOD__
);
5116 $dbr = wfGetDB( DB_SLAVE
);
5117 $timestamp = $dbr->selectField( 'revision', 'rev_timestamp',
5118 array( 'rev_id' => $this->mRevisionId
), __METHOD__
);
5120 // Normalize timestamp to internal MW format for timezone processing.
5121 // This has the added side-effect of replacing a null value with
5122 // the current time, which gives us more sensible behavior for
5124 $timestamp = wfTimestamp( TS_MW
, $timestamp );
5126 // The cryptic '' timezone parameter tells to use the site-default
5127 // timezone offset instead of the user settings.
5129 // Since this value will be saved into the parser cache, served
5130 // to other users, and potentially even used inside links and such,
5131 // it needs to be consistent for all visitors.
5132 $this->mRevisionTimestamp
= $wgContLang->userAdjust( $timestamp, '' );
5134 wfProfileOut( __METHOD__
);
5136 return $this->mRevisionTimestamp
;
5140 * Mutator for $mDefaultSort
5142 * @param $sort New value
5144 public function setDefaultSort( $sort ) {
5145 $this->mDefaultSort
= $sort;
5149 * Accessor for $mDefaultSort
5150 * Will use the title/prefixed title if none is set
5154 public function getDefaultSort() {
5155 if( $this->mDefaultSort
!== false ) {
5156 return $this->mDefaultSort
;
5158 return $this->mTitle
->getNamespace() == NS_CATEGORY
5159 ?
$this->mTitle
->getText()
5160 : $this->mTitle
->getPrefixedText();
5165 * Try to guess the section anchor name based on a wikitext fragment
5166 * presumably extracted from a heading, for example "Header" from
5169 public function guessSectionNameFromWikiText( $text ) {
5170 # Strip out wikitext links(they break the anchor)
5171 $text = $this->stripSectionName( $text );
5172 $headline = Sanitizer
::decodeCharReferences( $text );
5174 $headline = StringUtils
::delimiterReplace( '<', '>', '', $headline );
5175 $headline = trim( $headline );
5176 $sectionanchor = '#' . urlencode( str_replace( ' ', '_', $headline ) );
5177 $replacearray = array(
5182 array_keys( $replacearray ),
5183 array_values( $replacearray ),
5188 * Strips a text string of wikitext for use in a section anchor
5190 * Accepts a text string and then removes all wikitext from the
5191 * string and leaves only the resultant text (i.e. the result of
5192 * [[User:WikiSysop|Sysop]] would be "Sysop" and the result of
5193 * [[User:WikiSysop]] would be "User:WikiSysop") - this is intended
5194 * to create valid section anchors by mimicing the output of the
5195 * parser when headings are parsed.
5197 * @param $text string Text string to be stripped of wikitext
5198 * for use in a Section anchor
5199 * @return Filtered text string
5201 public function stripSectionName( $text ) {
5202 # Strip internal link markup
5203 $text = preg_replace('/\[\[:?([^[|]+)\|([^[]+)\]\]/','$2',$text);
5204 $text = preg_replace('/\[\[:?([^[]+)\|?\]\]/','$1',$text);
5206 # Strip external link markup (FIXME: Not Tolerant to blank link text
5207 # I.E. [http://www.mediawiki.org] will render as [1] or something depending
5208 # on how many empty links there are on the page - need to figure that out.
5209 $text = preg_replace('/\[(?:' . wfUrlProtocols() . ')([^ ]+?) ([^[]+)\]/','$2',$text);
5211 # Parse wikitext quotes (italics & bold)
5212 $text = $this->doQuotes($text);
5215 $text = StringUtils
::delimiterReplace( '<', '>', '', $text );
5220 * strip/replaceVariables/unstrip for preprocessor regression testing
5222 function srvus( $text ) {
5223 $text = $this->replaceVariables( $text );
5224 $text = $this->mStripState
->unstripBoth( $text );
5225 $text = Sanitizer
::removeHTMLtags( $text );
5231 * @todo document, briefly.
5232 * @addtogroup Parser
5234 class OnlyIncludeReplacer
{
5237 function replace( $matches ) {
5238 if ( substr( $matches[1], -1 ) == "\n" ) {
5239 $this->output
.= substr( $matches[1], 0, -1 );
5241 $this->output
.= $matches[1];
5247 * @todo document, briefly.
5248 * @addtogroup Parser
5251 var $general, $nowiki;
5253 function __construct() {
5254 $this->general
= new ReplacementArray
;
5255 $this->nowiki
= new ReplacementArray
;
5258 function unstripGeneral( $text ) {
5259 wfProfileIn( __METHOD__
);
5262 $text = $this->general
->replace( $text );
5263 } while ( $text != $oldText );
5264 wfProfileOut( __METHOD__
);
5268 function unstripNoWiki( $text ) {
5269 wfProfileIn( __METHOD__
);
5272 $text = $this->nowiki
->replace( $text );
5273 } while ( $text != $oldText );
5274 wfProfileOut( __METHOD__
);
5278 function unstripBoth( $text ) {
5279 wfProfileIn( __METHOD__
);
5282 $text = $this->general
->replace( $text );
5283 $text = $this->nowiki
->replace( $text );
5284 } while ( $text != $oldText );
5285 wfProfileOut( __METHOD__
);
5291 * An expansion frame, used as a context to expand the result of preprocessToDom()
5294 var $parser, $title;
5298 const NO_TEMPLATES
= 2;
5299 const STRIP_COMMENTS
= 4;
5300 const NO_IGNORE
= 8;
5302 const RECOVER_ORIG
= 11;
5305 * Construct a new preprocessor frame.
5306 * @param Parser $parser The parent parser
5307 * @param Title $title The context title, or false if there isn't one
5309 function __construct( $parser ) {
5310 $this->parser
= $parser;
5311 $this->title
= $parser->mTitle
;
5312 $this->titleCache
= array( $this->title ?
$this->title
->getPrefixedDBkey() : false );
5316 * Create a new child frame
5317 * $args is optionally a DOMNodeList containing the template arguments
5319 function newChild( $args = false, $title = false ) {
5320 $assocArgs = array();
5321 if ( $title === false ) {
5322 $title = $this->title
;
5324 if ( $args !== false ) {
5326 foreach ( $args as $arg ) {
5328 $xpath = new DOMXPath( $arg->ownerDocument
);
5331 $nameNodes = $xpath->query( 'name', $arg );
5332 if ( $nameNodes->item( 0 )->hasAttributes() ) {
5333 // Numbered parameter
5334 $name = $nameNodes->item( 0 )->attributes
->getNamedItem( 'index' )->textContent
;
5337 $name = $this->expand( $nameNodes->item( 0 ), PPFrame
::STRIP_COMMENTS
);
5340 $value = $xpath->query( 'value', $arg );
5341 $assocArgs[$name] = $value->item( 0 );
5344 return new PPTemplateFrame( $this->parser
, $this, $assocArgs, $title );
5348 * Expand a DOMNode describing a preprocessed document into plain wikitext,
5349 * using the current context
5350 * @param $root the node
5352 function expand( $root, $flags = 0 ) {
5353 if ( is_string( $root ) ) {
5357 if ( $this->parser
->ot
['html']
5358 && ++
$this->parser
->mPPNodeCount
> $this->parser
->mOptions
->mMaxPPNodeCount
)
5360 return $this->parser
->insertStripItem( '<!-- node-count limit exceeded -->' );
5363 if ( is_array( $root ) ) {
5365 foreach ( $root as $node ) {
5366 $s .= $this->expand( $node, $flags );
5368 } elseif ( $root instanceof DOMNodeList
) {
5370 foreach ( $root as $node ) {
5371 $s .= $this->expand( $node, $flags );
5373 } elseif ( $root instanceof DOMNode
) {
5374 if ( $root->nodeType
== XML_TEXT_NODE
) {
5375 $s = $root->nodeValue
;
5376 } elseif ( $root->nodeName
== 'template' ) {
5377 # Double-brace expansion
5378 $xpath = new DOMXPath( $root->ownerDocument
);
5379 $titles = $xpath->query( 'title', $root );
5380 $title = $titles->item( 0 );
5381 $parts = $xpath->query( 'part', $root );
5382 if ( $flags & self
::NO_TEMPLATES
) {
5383 $s = '{{' . $this->implodeWithFlags( '|', $flags, $title, $parts ) . '}}';
5385 $lineStart = $root->getAttribute( 'lineStart' );
5389 'lineStart' => $lineStart,
5390 'text' => 'FIXME' );
5391 $s = $this->parser
->braceSubstitution( $params, $this );
5393 } elseif ( $root->nodeName
== 'tplarg' ) {
5394 # Triple-brace expansion
5395 $xpath = new DOMXPath( $root->ownerDocument
);
5396 $titles = $xpath->query( 'title', $root );
5397 $title = $titles->item( 0 );
5398 $parts = $xpath->query( 'part', $root );
5399 if ( $flags & self
::NO_ARGS ||
$this->parser
->ot
['msg'] ) {
5400 $s = '{{{' . $this->implodeWithFlags( '|', $flags, $title, $parts ) . '}}}';
5402 $params = array( 'title' => $title, 'parts' => $parts, 'text' => 'FIXME' );
5403 $s = $this->parser
->argSubstitution( $params, $this );
5405 } elseif ( $root->nodeName
== 'comment' ) {
5406 # HTML-style comment
5407 if ( $this->parser
->ot
['html']
5408 ||
( $this->parser
->ot
['pre'] && $this->mOptions
->getRemoveComments() )
5409 ||
( $flags & self
::STRIP_COMMENTS
) )
5413 $s = $root->textContent
;
5415 } elseif ( $root->nodeName
== 'ignore' ) {
5416 # Output suppression used by <includeonly> etc.
5417 # OT_WIKI will only respect <ignore> in substed templates.
5418 # The other output types respect it unless NO_IGNORE is set.
5419 # extractSections() sets NO_IGNORE and so never respects it.
5420 if ( ( !isset( $this->parent
) && $this->parser
->ot
['wiki'] ) ||
( $flags & self
::NO_IGNORE
) ) {
5421 $s = $root->textContent
;
5425 } elseif ( $root->nodeName
== 'ext' ) {
5427 $xpath = new DOMXPath( $root->ownerDocument
);
5428 $names = $xpath->query( 'name', $root );
5429 $attrs = $xpath->query( 'attr', $root );
5430 $inners = $xpath->query( 'inner', $root );
5431 $closes = $xpath->query( 'close', $root );
5433 'name' => $names->item( 0 ),
5434 'attr' => $attrs->length
> 0 ?
$attrs->item( 0 ) : null,
5435 'inner' => $inners->length
> 0 ?
$inners->item( 0 ) : null,
5436 'close' => $closes->length
> 0 ?
$closes->item( 0 ) : null,
5438 $s = $this->parser
->extensionSubstitution( $params, $this );
5439 } elseif ( $root->nodeName
== 'h' ) {
5441 $s = $this->expand( $root->childNodes
, $flags );
5443 if ( $this->parser
->ot
['html'] ) {
5444 # Insert heading index marker
5445 $headingIndex = $root->getAttribute( 'i' );
5446 $titleText = $this->title
->getPrefixedDBkey();
5447 $this->parser
->mHeadings
[] = array( $titleText, $headingIndex );
5448 $serial = count( $this->parser
->mHeadings
) - 1;
5449 $marker = "{$this->parser->mUniqPrefix}-h-$serial-{$this->parser->mMarkerSuffix}";
5450 $count = $root->getAttribute( 'level' );
5451 $s = substr( $s, 0, $count ) . $marker . substr( $s, $count );
5452 $this->parser
->mStripState
->general
->setPair( $marker, '' );
5455 # Generic recursive expansion
5457 for ( $node = $root->firstChild
; $node; $node = $node->nextSibling
) {
5458 if ( $node->nodeType
== XML_TEXT_NODE
) {
5459 $s .= $node->nodeValue
;
5460 } elseif ( $node->nodeType
== XML_ELEMENT_NODE
) {
5461 $s .= $this->expand( $node, $flags );
5466 throw new MWException( __METHOD__
.': Invalid parameter type' );
5471 function implodeWithFlags( $sep, $flags /*, ... */ ) {
5472 $args = array_slice( func_get_args(), 2 );
5476 foreach ( $args as $root ) {
5477 if ( !is_array( $root ) && !( $root instanceof DOMNodeList
) ) {
5478 $root = array( $root );
5480 foreach ( $root as $node ) {
5486 $s .= $this->expand( $node, $flags );
5492 function implode( $sep /*, ... */ ) {
5493 $args = func_get_args();
5494 $args = array_merge( array_slice( $args, 0, 1 ), array( 0 ), array_slice( $args, 1 ) );
5495 return call_user_func_array( array( $this, 'implodeWithFlags' ), $args );
5499 * Split an <arg> or <template> node into a three-element array:
5500 * DOMNode name, string index and DOMNode value
5502 function splitBraceNode( $node ) {
5503 $xpath = new DOMXPath( $node->ownerDocument
);
5504 $names = $xpath->query( 'name', $node );
5505 $values = $xpath->query( 'value', $node );
5506 if ( !$names->length ||
!$values->length
) {
5507 throw new MWException( 'Invalid brace node passed to ' . __METHOD__
);
5509 $name = $names->item( 0 );
5510 $index = $name->getAttribute( 'index' );
5511 return array( $name, $index, $values->item( 0 ) );
5515 * Split an <ext> node into an associative array containing name, attr, inner and close
5516 * All values in the resulting array are DOMNodes. Inner and close are optional.
5518 function splitExtNode( $node ) {
5519 $xpath = new DOMXPath( $node->ownerDocument
);
5520 $names = $xpath->query( 'name', $node );
5521 $attrs = $xpath->query( 'attr', $node );
5522 $inners = $xpath->query( 'inner', $node );
5523 $closes = $xpath->query( 'close', $node );
5524 if ( !$names->length ||
!$attrs->length
) {
5525 throw new MWException( 'Invalid ext node passed to ' . __METHOD__
);
5528 'name' => $names->item( 0 ),
5529 'attr' => $attrs->item( 0 ) );
5530 if ( $inners->length
) {
5531 $parts['inner'] = $inners->item( 0 );
5533 if ( $closes->length
) {
5534 $parts['close'] = $closes->item( 0 );
5539 function __toString() {
5543 function getPDBK( $level = false ) {
5544 if ( $level === false ) {
5545 return $this->title
->getPrefixedDBkey();
5547 return isset( $this->titleCache
[$level] ) ?
$this->titleCache
[$level] : false;
5553 * Expansion frame with template arguments
5555 class PPTemplateFrame
extends PPFrame
{
5556 var $parser, $args, $parent;
5559 function __construct( $parser, $parent = false, $args = array(), $title = false ) {
5560 $this->parser
= $parser;
5561 $this->parent
= $parent;
5562 $this->args
= $args;
5563 $this->title
= $title;
5564 $this->titleCache
= $parent->titleCache
;
5565 $this->titleCache
[] = $title ?
$title->getPrefixedDBkey() : false;
5568 function __toString() {
5571 foreach ( $this->args
as $name => $value ) {
5577 $s .= "\"$name\":\"" .
5578 str_replace( '"', '\\"', $value->ownerDocument
->saveXML( $value ) ) . '"';