25 use Wikimedia\ScopedCallback;
76 const VERSION =
'1.6.4';
82 const HALF_PARSED_VERSION = 2;
84 # Flags for Parser::setFunctionHook
88 # Constants needed for external link processing
89 # Everything except bracket, space, or control characters
90 # \p{Zs} is unicode 'separator, space' category. It covers the space 0x20
91 # as well as U+3000 is IDEOGRAPHIC SPACE for T21052
92 # \x{FFFD} is the Unicode replacement character, which Preprocessor_DOM
93 # uses to replace invalid HTML characters.
94 const EXT_LINK_URL_CLASS =
'[^][<>"\\x00-\\x20\\x7F\p{Zs}\x{FFFD}]';
95 # Simplified expression to match an IPv4 or IPv6 address, or
96 # at least one character of a host name (embeds EXT_LINK_URL_CLASS)
97 const EXT_LINK_ADDR =
'(?:[0-9.]+|\\[(?i:[0-9a-f:.]+)\\]|[^][<>"\\x00-\\x20\\x7F\p{Zs}\x{FFFD}])';
98 # RegExp to make image URLs (embeds IPv6 part of EXT_LINK_ADDR)
100 const EXT_IMAGE_REGEX =
'/^(http:\/\/|https:\/\/)((?:\\[(?i:[0-9a-f:.]+)\\])?[^][<>"\\x00-\\x20\\x7F\p{Zs}\x{FFFD}]+)
101 \\/([A-Za-z0-9_.,~%\\-+&;#*?!=()@\\x80-\\xFF]+)\\.((?i)gif|png|jpg|jpeg)$/Sxu';
104 # Regular expression for a non-newline space
105 const SPACE_NOT_NL =
'(?:\t| |&\#0*160;|&\#[Xx]0*[Aa]0;|\p{Zs})';
107 # Flags for preprocessToDom
108 const PTD_FOR_INCLUSION = 1;
110 # Allowed values for $this->mOutputType
111 # Parameter to startExternalParse().
113 const
OT_WIKI = 2;
# like preSaveTransform()
116 const
OT_PLAIN = 4;
# like extractSections() - portions of the original are returned unchanged.
135 const MARKER_SUFFIX =
"-QINU`\"'\x7f";
136 const MARKER_PREFIX =
"\x7f'\"`UNIQ-";
138 # Markers used for wrapping the table of contents
139 const TOC_START =
'<mw:toc>';
140 const TOC_END =
'</mw:toc>';
143 public $mTagHooks = [];
144 public $mTransparentTagHooks = [];
145 public $mFunctionHooks = [];
146 public $mFunctionSynonyms = [ 0 => [], 1 => [] ];
147 public $mFunctionTagHooks = [];
148 public $mStripList = [];
149 public $mDefaultStripList = [];
150 public $mVarCache = [];
151 public $mImageParams = [];
152 public $mImageParamsMagicArray = [];
153 public $mMarkerIndex = 0;
154 public $mFirstCall =
true;
156 # Initialised by initialiseVariables()
167 # Initialised in constructor
168 public $mConf, $mExtLinkBracketedRegex, $mUrlProtocols;
170 # Initialized in getPreprocessor()
172 public $mPreprocessor;
174 # Cleared with clearState():
186 public $mIncludeCount;
190 public $mLinkHolders;
193 public $mIncludeSizes, $mPPNodeCount, $mGeneratedPPNodeCount, $mHighestExpansionDepth;
194 public $mDefaultSort;
195 public $mTplRedirCache, $mTplDomCache, $mHeadings, $mDoubleUnderscores;
196 public $mExpensiveFunctionCount; # number
of expensive
parser function calls
197 public $mShowToc, $mForceTocPosition;
205 # These are variables reset at least once per parse regardless of $clearState
216 public $mOutputType; # Output
type, one
of the OT_xxx constants
217 public $ot; # Shortcut alias,
see setOutputType()
218 public $mRevisionObject;
# The revision object of the specified revision ID
219 public $mRevisionId; # ID to
display in {{REVISIONID}} tags
220 public $mRevisionTimestamp; # The timestamp
of the
specified revision ID
223 public $mRevIdForTs; # The revision ID which was
used to fetch the timestamp
224 public $mInputSize =
false; # For {{PAGESIZE}}
on current
page.
230 public $mUniqPrefix = self::MARKER_PREFIX;
237 public $mLangLinkLanguages;
245 public $currentRevisionCache;
251 public $mInParse =
false;
254 protected $mProfiler;
259 protected $mLinkRenderer;
264 public function __construct( $conf = [] ) {
265 $this->mConf = $conf;
267 $this->mExtLinkBracketedRegex =
'/\[(((?i)' . $this->mUrlProtocols .
')' .
268 self::EXT_LINK_ADDR .
269 self::EXT_LINK_URL_CLASS .
'*)\p{Zs}*([^\]\\x00-\\x08\\x0a-\\x1F\\x{FFFD}]*?)\]/Su';
270 if ( isset( $conf[
'preprocessorClass'] ) ) {
271 $this->mPreprocessorClass = $conf[
'preprocessorClass'];
272 } elseif ( defined(
'HPHP_VERSION' ) ) {
273 # Preprocessor_Hash is much faster than Preprocessor_DOM under HipHop
274 $this->mPreprocessorClass =
'Preprocessor_Hash';
275 } elseif ( extension_loaded(
'domxml' ) ) {
276 # PECL extension that conflicts with the core DOM extension (T15770)
277 wfDebug(
"Warning: you have the obsolete domxml extension for PHP. Please remove it!\n" );
278 $this->mPreprocessorClass =
'Preprocessor_Hash';
279 } elseif ( extension_loaded(
'dom' ) ) {
280 $this->mPreprocessorClass =
'Preprocessor_DOM';
282 $this->mPreprocessorClass =
'Preprocessor_Hash';
284 wfDebug( __CLASS__ .
": using preprocessor: {$this->mPreprocessorClass}\n" );
290 public function __destruct() {
291 if ( isset( $this->mLinkHolders ) ) {
292 unset( $this->mLinkHolders );
295 unset( $this->$name );
302 public function __clone() {
303 $this->mInParse =
false;
311 foreach ( [
'mStripState',
'mVarCache' ]
as $k ) {
325 public function firstCallInit() {
326 if ( !$this->mFirstCall ) {
329 $this->mFirstCall =
false;
333 $this->initialiseVariables();
345 public function clearState() {
346 if ( $this->mFirstCall ) {
347 $this->firstCallInit();
350 $this->mOptions->registerWatcher( [ $this->mOutput,
'recordOption' ] );
351 $this->mAutonumber = 0;
352 $this->mIncludeCount = [];
355 $this->mRevisionObject = $this->mRevisionTimestamp =
356 $this->mRevisionId = $this->mRevisionUser = $this->mRevisionSize =
null;
357 $this->mVarCache = [];
359 $this->mLangLinkLanguages = [];
360 $this->currentRevisionCache =
null;
364 # Clear these on every parse, T6549
365 $this->mTplRedirCache = $this->mTplDomCache = [];
367 $this->mShowToc =
true;
368 $this->mForceTocPosition =
false;
369 $this->mIncludeSizes = [
373 $this->mPPNodeCount = 0;
374 $this->mGeneratedPPNodeCount = 0;
375 $this->mHighestExpansionDepth = 0;
376 $this->mDefaultSort =
false;
377 $this->mHeadings = [];
378 $this->mDoubleUnderscores = [];
379 $this->mExpensiveFunctionCount = 0;
382 if ( isset( $this->mPreprocessor ) && $this->mPreprocessor->parser !== $this ) {
383 $this->mPreprocessor =
null;
405 public function parse(
407 $linestart =
true, $clearState =
true, $revid =
null
419 $text = strtr( $text,
"\x7f",
"?" );
420 $magicScopeVariable = $this->lock();
423 $text = str_replace(
"\000",
'', $text );
427 $this->currentRevisionCache =
null;
428 $this->mInputSize = strlen( $text );
429 if ( $this->mOptions->getEnableLimitReport() ) {
430 $this->mOutput->resetParseStartTime();
433 $oldRevisionId = $this->mRevisionId;
434 $oldRevisionObject = $this->mRevisionObject;
435 $oldRevisionTimestamp = $this->mRevisionTimestamp;
436 $oldRevisionUser = $this->mRevisionUser;
437 $oldRevisionSize = $this->mRevisionSize;
438 if ( $revid !==
null ) {
439 $this->mRevisionId = $revid;
440 $this->mRevisionObject =
null;
441 $this->mRevisionTimestamp =
null;
442 $this->mRevisionUser =
null;
443 $this->mRevisionSize =
null;
451 $text = $this->internalParse( $text );
454 $text = $this->internalParseHalfParsed( $text,
true, $linestart );
463 if ( !(
$options->getDisableTitleConversion()
464 || isset( $this->mDoubleUnderscores[
'nocontentconvert'] )
465 || isset( $this->mDoubleUnderscores[
'notitleconvert'] )
466 || $this->mOutput->getDisplayTitle() !==
false )
468 $convruletitle = $this->getConverterLanguage()->getConvRuleTitle();
469 if ( $convruletitle ) {
470 $this->mOutput->setTitleText( $convruletitle );
472 $titleText = $this->getConverterLanguage()->convertTitle(
$title );
473 $this->mOutput->setTitleText( $titleText );
477 # Done parsing! Compute runtime adaptive expiry if set
478 $this->mOutput->finalizeAdaptiveCacheExpiry();
480 # Warn if too many heavyweight parser functions were used
481 if ( $this->mExpensiveFunctionCount > $this->mOptions->getExpensiveParserFunctionLimit() ) {
482 $this->limitationWarn(
'expensive-parserfunction',
483 $this->mExpensiveFunctionCount,
484 $this->mOptions->getExpensiveParserFunctionLimit()
488 # Information on include size limits, for the benefit of users who try to skirt them
489 if ( $this->mOptions->getEnableLimitReport() ) {
490 $max = $this->mOptions->getMaxIncludeSize();
492 $cpuTime = $this->mOutput->getTimeSinceStart(
'cpu' );
493 if ( $cpuTime !==
null ) {
494 $this->mOutput->setLimitReportData(
'limitreport-cputime',
495 sprintf(
"%.3f", $cpuTime )
499 $wallTime = $this->mOutput->getTimeSinceStart(
'wall' );
500 $this->mOutput->setLimitReportData(
'limitreport-walltime',
501 sprintf(
"%.3f", $wallTime )
504 $this->mOutput->setLimitReportData(
'limitreport-ppvisitednodes',
505 [ $this->mPPNodeCount, $this->mOptions->getMaxPPNodeCount() ]
507 $this->mOutput->setLimitReportData(
'limitreport-ppgeneratednodes',
508 [ $this->mGeneratedPPNodeCount, $this->mOptions->getMaxGeneratedPPNodeCount() ]
510 $this->mOutput->setLimitReportData(
'limitreport-postexpandincludesize',
511 [ $this->mIncludeSizes[
'post-expand'], $max ]
513 $this->mOutput->setLimitReportData(
'limitreport-templateargumentsize',
514 [ $this->mIncludeSizes[
'arg'], $max ]
516 $this->mOutput->setLimitReportData(
'limitreport-expansiondepth',
517 [ $this->mHighestExpansionDepth, $this->mOptions->getMaxPPExpandDepth() ]
519 $this->mOutput->setLimitReportData(
'limitreport-expensivefunctioncount',
520 [ $this->mExpensiveFunctionCount, $this->mOptions->getExpensiveParserFunctionLimit() ]
522 Hooks::run(
'ParserLimitReportPrepare', [ $this, $this->mOutput ] );
524 $limitReport =
"NewPP limit report\n";
526 $limitReport .=
'Parsed by ' .
wfHostname() .
"\n";
528 $limitReport .=
'Cached time: ' . $this->mOutput->getCacheTime() .
"\n";
529 $limitReport .=
'Cache expiry: ' . $this->mOutput->getCacheExpiry() .
"\n";
530 $limitReport .=
'Dynamic content: ' .
531 ( $this->mOutput->hasDynamicContent() ?
'true' :
'false' ) .
534 foreach ( $this->mOutput->getLimitReportData()
as $key =>
$value ) {
536 [ $key, &
$value, &$limitReport,
false,
false ]
538 $keyMsg =
wfMessage( $key )->inLanguage(
'en' )->useDatabase(
false );
539 $valueMsg =
wfMessage( [
"$key-value-text",
"$key-value" ] )
540 ->inLanguage(
'en' )->useDatabase(
false );
541 if ( !$valueMsg->exists() ) {
544 if ( !$keyMsg->isDisabled() && !$valueMsg->isDisabled() ) {
545 $valueMsg->params(
$value );
546 $limitReport .=
"{$keyMsg->text()}: {$valueMsg->text()}\n";
552 $limitReport = htmlspecialchars_decode( $limitReport );
554 Hooks::run(
'ParserLimitReport', [ $this, &$limitReport ],
'1.22' );
558 $limitReport = str_replace( [
'-',
'&' ], [
'‐',
'&' ], $limitReport );
559 $text .=
"\n<!-- \n$limitReport-->\n";
562 $dataByFunc = $this->mProfiler->getFunctionStats();
563 uasort( $dataByFunc,
function ( $a, $b ) {
564 return $a[
'real'] < $b[
'real'];
567 foreach ( array_slice( $dataByFunc, 0, 10 )
as $item ) {
568 $profileReport[] = sprintf(
"%6.2f%% %8.3f %6d %s",
569 $item[
'%real'], $item[
'real'], $item[
'calls'],
570 htmlspecialchars( $item[
'name'] ) );
572 $text .=
"<!--\nTransclusion expansion time report (%,ms,calls,template)\n";
573 $text .= implode(
"\n", $profileReport ) .
"\n-->\n";
575 $this->mOutput->setLimitReportData(
'limitreport-timingprofile', $profileReport );
579 $this->mOutput->setLimitReportData(
'cachereport-origin',
wfHostname() );
581 $this->mOutput->setLimitReportData(
'cachereport-timestamp',
582 $this->mOutput->getCacheTime() );
583 $this->mOutput->setLimitReportData(
'cachereport-ttl',
584 $this->mOutput->getCacheExpiry() );
585 $this->mOutput->setLimitReportData(
'cachereport-transientcontent',
586 $this->mOutput->hasDynamicContent() );
588 if ( $this->mGeneratedPPNodeCount > $this->mOptions->getMaxGeneratedPPNodeCount() / 10 ) {
589 wfDebugLog(
'generated-pp-node-count', $this->mGeneratedPPNodeCount .
' ' .
590 $this->mTitle->getPrefixedDBkey() );
594 # Wrap non-interface parser output in a <div> so it can be targeted
596 $class = $this->mOptions->getWrapOutputClass();
597 if ( $class !==
false && !$this->mOptions->getInterfaceMessage() ) {
601 $this->mOutput->setText( $text );
603 $this->mRevisionId = $oldRevisionId;
604 $this->mRevisionObject = $oldRevisionObject;
605 $this->mRevisionTimestamp = $oldRevisionTimestamp;
606 $this->mRevisionUser = $oldRevisionUser;
607 $this->mRevisionSize = $oldRevisionSize;
608 $this->mInputSize =
false;
609 $this->currentRevisionCache =
null;
611 return $this->mOutput;
636 public function recursiveTagParse( $text, $frame =
false ) {
641 $text = $this->internalParse( $text,
false, $frame );
662 public function recursiveTagParseFully( $text, $frame =
false ) {
663 $text = $this->recursiveTagParse( $text, $frame );
664 $text = $this->internalParseHalfParsed( $text,
false );
679 public function preprocess( $text,
Title $title =
null,
682 $magicScopeVariable = $this->lock();
684 if ( $revid !==
null ) {
685 $this->mRevisionId = $revid;
691 $text = $this->replaceVariables( $text, $frame );
692 $text = $this->mStripState->unstripBoth( $text );
705 public function recursivePreprocess( $text, $frame =
false ) {
706 $text = $this->replaceVariables( $text, $frame );
707 $text = $this->mStripState->unstripBoth( $text );
726 $text = $msg->params(
$params )->plain();
728 # Parser (re)initialisation
729 $magicScopeVariable = $this->lock();
733 $dom = $this->preprocessToDom( $text, self::PTD_FOR_INCLUSION );
734 $text = $this->getPreprocessor()->newFrame()->expand( $dom,
$flags );
735 $text = $this->mStripState->unstripBoth( $text );
745 public function setUser(
$user ) {
746 $this->mUser =
$user;
754 public function setTitle(
$t ) {
759 if (
$t->hasFragment() ) {
760 # Strip the fragment to avoid various odd effects
761 $this->mTitle =
$t->createFragmentTarget(
'' );
772 public function getTitle() {
773 return $this->mTitle;
782 public function Title( $x =
null ) {
783 return wfSetVar( $this->mTitle, $x );
791 public function setOutputType( $ot ) {
792 $this->mOutputType = $ot;
808 public function OutputType( $x =
null ) {
809 return wfSetVar( $this->mOutputType, $x );
817 public function getOutput() {
818 return $this->mOutput;
826 public function getOptions() {
827 return $this->mOptions;
836 public function Options( $x =
null ) {
837 return wfSetVar( $this->mOptions, $x );
843 public function nextLinkID() {
844 return $this->mLinkID++;
850 public function setLinkID( $id ) {
851 $this->mLinkID = $id;
858 public function getFunctionLang() {
859 return $this->getTargetLanguage();
871 public function getTargetLanguage() {
872 $target = $this->mOptions->getTargetLanguage();
874 if ( $target !==
null ) {
876 } elseif ( $this->mOptions->getInterfaceMessage() ) {
877 return $this->mOptions->getUserLangObj();
878 } elseif ( is_null( $this->mTitle ) ) {
879 throw new MWException( __METHOD__ .
': $this->mTitle is null' );
882 return $this->mTitle->getPageLanguage();
889 public function getConverterLanguage() {
890 return $this->getTargetLanguage();
899 public function getUser() {
900 if ( !is_null( $this->mUser ) ) {
903 return $this->mOptions->getUser();
911 public function getPreprocessor() {
912 if ( !isset( $this->mPreprocessor ) ) {
913 $class = $this->mPreprocessorClass;
914 $this->mPreprocessor =
new $class( $this );
916 return $this->mPreprocessor;
925 public function getLinkRenderer() {
926 if ( !$this->mLinkRenderer ) {
927 $this->mLinkRenderer = MediaWikiServices::getInstance()
928 ->getLinkRendererFactory()->create();
929 $this->mLinkRenderer->setStubThreshold(
930 $this->getOptions()->getStubThreshold()
934 return $this->mLinkRenderer;
956 public static function extractTagsAndParams( $elements, $text, &
$matches ) {
961 $taglist = implode(
'|', $elements );
962 $start =
"/<($taglist)(\\s+[^>]*?|\\s*?)(\/?" .
">)|<(!--)/i";
964 while ( $text !=
'' ) {
965 $p = preg_split( $start, $text, 2, PREG_SPLIT_DELIM_CAPTURE );
967 if (
count( $p ) < 5 ) {
970 if (
count( $p ) > 5 ) {
984 $marker = self::MARKER_PREFIX .
"-$element-" . sprintf(
'%08X', $n++ ) . self::MARKER_SUFFIX;
985 $stripped .= $marker;
987 if ( $close ===
'/>' ) {
988 # Empty element tag, <tag />
993 if ( $element ===
'!--' ) {
996 $end =
"/(<\\/$element\\s*>)/i";
998 $q = preg_split( $end, $inside, 2, PREG_SPLIT_DELIM_CAPTURE );
1000 if (
count( $q ) < 3 ) {
1001 # No end tag -- let it run out to the end of the text.
1012 Sanitizer::decodeTagAttributes( $attributes ),
1013 "<$element$attributes$close$content$tail" ];
1023 public function getStripList() {
1024 return $this->mStripList;
1036 public function insertStripItem( $text ) {
1037 $marker = self::MARKER_PREFIX .
"-item-{$this->mMarkerIndex}-" . self::MARKER_SUFFIX;
1038 $this->mMarkerIndex++;
1039 $this->mStripState->addGeneral( $marker, $text );
1050 public function doTableStuff( $text ) {
1053 $td_history = []; # Is currently
a td
tag open?
1054 $last_tag_history = []; # Save
history of last lag activated (td, th
or caption)
1055 $tr_history = []; # Is currently
a tr
tag open?
1056 $tr_attributes = []; #
history of tr attributes
1057 $has_opened_tr = []; # Did
this table open a <tr> element?
1058 $indent_level = 0; # indent level
of the
table
1061 $line = trim( $outLine );
1064 $out .= $outLine .
"\n";
1068 $first_character =
$line[0];
1069 $first_two = substr(
$line, 0, 2 );
1072 if ( preg_match(
'/^(:*)\s*\{\|(.*)$/',
$line,
$matches ) ) {
1073 # First check if we are starting a new table
1074 $indent_level = strlen(
$matches[1] );
1076 $attributes = $this->mStripState->unstripBoth(
$matches[2] );
1077 $attributes = Sanitizer::fixTagAttributes( $attributes,
'table' );
1079 $outLine = str_repeat(
'<dl><dd>', $indent_level ) .
"<table{$attributes}>";
1080 array_push( $td_history,
false );
1081 array_push( $last_tag_history,
'' );
1082 array_push( $tr_history,
false );
1083 array_push( $tr_attributes,
'' );
1084 array_push( $has_opened_tr,
false );
1085 } elseif (
count( $td_history ) == 0 ) {
1086 # Don't do any of the following
1087 $out .= $outLine .
"\n";
1089 } elseif ( $first_two ===
'|}' ) {
1090 # We are ending a table
1092 $last_tag = array_pop( $last_tag_history );
1094 if ( !array_pop( $has_opened_tr ) ) {
1095 $line =
"<tr><td></td></tr>{$line}";
1098 if ( array_pop( $tr_history ) ) {
1099 $line =
"</tr>{$line}";
1102 if ( array_pop( $td_history ) ) {
1103 $line =
"</{$last_tag}>{$line}";
1105 array_pop( $tr_attributes );
1106 $outLine =
$line . str_repeat(
'</dd></dl>', $indent_level );
1107 } elseif ( $first_two ===
'|-' ) {
1108 # Now we have a table row
1109 $line = preg_replace(
'#^\|-+#',
'',
$line );
1111 # Whats after the tag is now only attributes
1112 $attributes = $this->mStripState->unstripBoth(
$line );
1113 $attributes = Sanitizer::fixTagAttributes( $attributes,
'tr' );
1114 array_pop( $tr_attributes );
1115 array_push( $tr_attributes, $attributes );
1118 $last_tag = array_pop( $last_tag_history );
1119 array_pop( $has_opened_tr );
1120 array_push( $has_opened_tr,
true );
1122 if ( array_pop( $tr_history ) ) {
1126 if ( array_pop( $td_history ) ) {
1127 $line =
"</{$last_tag}>{$line}";
1131 array_push( $tr_history,
false );
1132 array_push( $td_history,
false );
1133 array_push( $last_tag_history,
'' );
1134 } elseif ( $first_character ===
'|'
1135 || $first_character ===
'!'
1136 || $first_two ===
'|+'
1138 # This might be cell elements, td, th or captions
1139 if ( $first_two ===
'|+' ) {
1140 $first_character =
'+';
1147 if ( $first_character ===
'!' ) {
1151 # Split up multiple cells on the same line.
1152 # FIXME : This can result in improper nesting of tags processed
1153 # by earlier parser steps.
1154 $cells = explode(
'||',
$line );
1158 # Loop through each table cell
1159 foreach ( $cells
as $cell ) {
1161 if ( $first_character !==
'+' ) {
1162 $tr_after = array_pop( $tr_attributes );
1163 if ( !array_pop( $tr_history ) ) {
1164 $previous =
"<tr{$tr_after}>\n";
1166 array_push( $tr_history,
true );
1167 array_push( $tr_attributes,
'' );
1168 array_pop( $has_opened_tr );
1169 array_push( $has_opened_tr,
true );
1172 $last_tag = array_pop( $last_tag_history );
1174 if ( array_pop( $td_history ) ) {
1175 $previous =
"</{$last_tag}>\n{$previous}";
1178 if ( $first_character ===
'|' ) {
1180 } elseif ( $first_character ===
'!' ) {
1182 } elseif ( $first_character ===
'+' ) {
1183 $last_tag =
'caption';
1188 array_push( $last_tag_history, $last_tag );
1190 # A cell could contain both parameters and data
1191 $cell_data = explode(
'|', $cell, 2 );
1193 # T2553: Note that a '|' inside an invalid link should not
1194 # be mistaken as delimiting cell parameters
1195 # Bug T153140: Neither should language converter markup.
1196 if ( preg_match(
'/\[\[|-\{/', $cell_data[0] ) === 1 ) {
1197 $cell =
"{$previous}<{$last_tag}>{$cell}";
1198 } elseif (
count( $cell_data ) == 1 ) {
1199 $cell =
"{$previous}<{$last_tag}>{$cell_data[0]}";
1201 $attributes = $this->mStripState->unstripBoth( $cell_data[0] );
1202 $attributes = Sanitizer::fixTagAttributes( $attributes, $last_tag );
1203 $cell =
"{$previous}<{$last_tag}{$attributes}>{$cell_data[1]}";
1207 array_push( $td_history,
true );
1210 $out .= $outLine .
"\n";
1213 # Closing open td, tr && table
1214 while (
count( $td_history ) > 0 ) {
1215 if ( array_pop( $td_history ) ) {
1218 if ( array_pop( $tr_history ) ) {
1221 if ( !array_pop( $has_opened_tr ) ) {
1222 $out .=
"<tr><td></td></tr>\n";
1225 $out .=
"</table>\n";
1228 # Remove trailing line-ending (b/c)
1229 if ( substr(
$out, -1 ) ===
"\n" ) {
1233 # special case: don't return empty table
1234 if (
$out ===
"<table>\n<tr><td></td></tr>\n</table>" ) {
1253 public function internalParse( $text, $isMain =
true, $frame =
false ) {
1259 # Hook to suspend the parser in this state
1260 if ( !
Hooks::run(
'ParserBeforeInternalParse', [ &
$parser, &$text, &$this->mStripState ] ) ) {
1264 # if $frame is provided, then use $frame for replacing any variables
1266 # use frame depth to infer how include/noinclude tags should be handled
1267 # depth=0 means this is the top-level document; otherwise it's an included document
1268 if ( !$frame->depth ) {
1271 $flag = self::PTD_FOR_INCLUSION;
1273 $dom = $this->preprocessToDom( $text, $flag );
1274 $text = $frame->expand( $dom );
1276 # if $frame is not provided, then use old-style replaceVariables
1277 $text = $this->replaceVariables( $text );
1280 Hooks::run(
'InternalParseBeforeSanitize', [ &
$parser, &$text, &$this->mStripState ] );
1281 $text = Sanitizer::removeHTMLtags(
1283 [ $this,
'attributeStripCallback' ],
1285 array_keys( $this->mTransparentTagHooks ),
1287 [ $this,
'addTrackingCategory' ]
1289 Hooks::run(
'InternalParseBeforeLinks', [ &
$parser, &$text, &$this->mStripState ] );
1291 # Tables need to come after variable replacement for things to work
1292 # properly; putting them before other transformations should keep
1293 # exciting things like link expansions from showing up in surprising
1295 $text = $this->doTableStuff( $text );
1297 $text = preg_replace(
'/(^|\n)-----*/',
'\\1<hr />', $text );
1299 $text = $this->doDoubleUnderscore( $text );
1301 $text = $this->doHeadings( $text );
1302 $text = $this->replaceInternalLinks( $text );
1303 $text = $this->doAllQuotes( $text );
1304 $text = $this->replaceExternalLinks( $text );
1306 # replaceInternalLinks may sometimes leave behind
1307 # absolute URLs, which have to be masked to hide them from replaceExternalLinks
1308 $text = str_replace( self::MARKER_PREFIX .
'NOPARSE',
'', $text );
1310 $text = $this->doMagicLinks( $text );
1311 $text = $this->formatHeadings( $text, $origText, $isMain );
1325 private function internalParseHalfParsed( $text, $isMain =
true, $linestart =
true ) {
1326 $text = $this->mStripState->unstripGeneral( $text );
1335 # Clean up special characters, only run once, next-to-last before doBlockLevels
1337 # French spaces, last one Guillemet-left
1338 # only if there is something before the space
1339 '/(.) (?=\\?|:|;|!|%|\\302\\273)/' =>
'\\1 ',
1340 # french spaces, Guillemet-right
1341 '/(\\302\\253) /' =>
'\\1 ',
1342 '/ (!\s*important)/' =>
' \\1', # Beware
of CSS magic word !important, T13874.
1344 $text = preg_replace( array_keys( $fixtags ), array_values( $fixtags ), $text );
1346 $text = $this->doBlockLevels( $text, $linestart );
1348 $this->replaceLinkHolders( $text );
1357 if ( !( $this->mOptions->getDisableContentConversion()
1358 || isset( $this->mDoubleUnderscores[
'nocontentconvert'] ) )
1360 if ( !$this->mOptions->getInterfaceMessage() ) {
1361 # The position of the convert() call should not be changed. it
1362 # assumes that the links are all replaced and the only thing left
1363 # is the <nowiki> mark.
1364 $text = $this->getConverterLanguage()->convert( $text );
1368 $text = $this->mStripState->unstripNoWiki( $text );
1374 $text = $this->replaceTransparentTags( $text );
1375 $text = $this->mStripState->unstripGeneral( $text );
1377 $text = Sanitizer::normalizeCharReferences( $text );
1380 if ( $this->mOptions->getTidy() ) {
1384 # attempt to sanitize at least some nesting problems
1385 # (T4702 and quite a few others)
1387 # ''Something [http://www.cool.com cool''] -->
1388 # <i>Something</i><a href="http://www.cool.com"..><i>cool></i></a>
1389 '/(<([bi])>)(<([bi])>)?([^<]*)(<\/?a[^<]*>)([^<]*)(<\/\\4>)?(<\/\\2>)/' =>
1390 '\\1\\3\\5\\8\\9\\6\\1\\3\\7\\8\\9',
1391 # fix up an anchor inside another anchor, only
1392 # at least for a single single nested link (T5695)
1393 '/(<a[^>]+>)([^<]*)(<a[^>]+>[^<]*)<\/a>(.*)<\/a>/' =>
1394 '\\1\\2</a>\\3</a>\\1\\4</a>',
1395 # fix div inside inline elements- doBlockLevels won't wrap a line which
1396 # contains a div, so fix it up here; replace
1397 # div with escaped text
1398 '/(<([aib]) [^>]+>)([^<]*)(<div([^>]*)>)(.*)(<\/div>)([^<]*)(<\/\\2>)/' =>
1399 '\\1\\3<div\\5>\\6</div>\\8\\9',
1400 # remove empty italic or bold tag pairs, some
1401 # introduced by rules above
1402 '/<([bi])><\/\\1>/' =>
'',
1405 $text = preg_replace(
1406 array_keys( $tidyregs ),
1407 array_values( $tidyregs ),
1429 public function doMagicLinks( $text ) {
1431 $urlChar = self::EXT_LINK_URL_CLASS;
1432 $addr = self::EXT_LINK_ADDR;
1433 $space = self::SPACE_NOT_NL; # non-newline space
1434 $spdash =
"(?:-|$space)"; #
a dash
or a non-newline space
1435 $spaces =
"$space++"; # possessive match
of 1
or more spaces
1436 $text = preg_replace_callback(
1438 (<a[ \t\r\n>].*?</a>) | # m[1]: Skip link text
1439 (<.*?>) | # m[2]: Skip stuff inside HTML elements' .
"
1440 (\b # m[3]: Free external links
1442 ($addr$urlChar*) # m[4]: Post-protocol path
1444 \b(?:RFC|PMID) $spaces # m[5]: RFC or PMID, capture number
1446 \bISBN $spaces ( # m[6]: ISBN, capture number
1447 (?: 97[89] $spdash? )? # optional 13-digit ISBN prefix
1448 (?: [0-9] $spdash? ){9} # 9 digits with opt. delimiters
1449 [0-9Xx] # check digit
1451 )!xu", [ $this,
'magicLinkCallback' ], $text );
1460 public function magicLinkCallback( $m ) {
1461 if ( isset( $m[1] ) && $m[1] !==
'' ) {
1464 } elseif ( isset( $m[2] ) && $m[2] !==
'' ) {
1467 } elseif ( isset( $m[3] ) && $m[3] !==
'' ) {
1468 # Free external link
1469 return $this->makeFreeExternalLink( $m[0], strlen( $m[4] ) );
1470 } elseif ( isset( $m[5] ) && $m[5] !==
'' ) {
1472 if ( substr( $m[0], 0, 3 ) ===
'RFC' ) {
1473 if ( !$this->mOptions->getMagicRFCLinks() ) {
1478 $cssClass =
'mw-magiclink-rfc';
1479 $trackingCat =
'magiclink-tracking-rfc';
1481 } elseif ( substr( $m[0], 0, 4 ) ===
'PMID' ) {
1482 if ( !$this->mOptions->getMagicPMIDLinks() ) {
1486 $urlmsg =
'pubmedurl';
1487 $cssClass =
'mw-magiclink-pmid';
1488 $trackingCat =
'magiclink-tracking-pmid';
1491 throw new MWException( __METHOD__ .
': unrecognised match type "' .
1492 substr( $m[0], 0, 20 ) .
'"' );
1494 $url =
wfMessage( $urlmsg, $id )->inContentLanguage()->text();
1495 $this->addTrackingCategory( $trackingCat );
1497 } elseif ( isset( $m[6] ) && $m[6] !==
''
1498 && $this->mOptions->getMagicISBNLinks()
1502 $space = self::SPACE_NOT_NL; # non-newline space
1503 $isbn = preg_replace(
"/$space/",
' ', $isbn );
1504 $num = strtr( $isbn, [
1509 $this->addTrackingCategory(
'magiclink-tracking-isbn' );
1510 return $this->getLinkRenderer()->makeKnownLink(
1514 'class' =>
'internal mw-magiclink-isbn',
1532 public function makeFreeExternalLink( $url, $numPostProto ) {
1535 # The characters '<' and '>' (which were escaped by
1536 # removeHTMLtags()) should not be included in
1537 # URLs, per RFC 2396.
1538 # Make terminate a URL as well (bug T84937)
1541 '/&(lt|gt|nbsp|#x0*(3[CcEe]|[Aa]0)|#0*(60|62|160));/',
1546 $trail = substr( $url, $m2[0][1] ) . $trail;
1547 $url = substr( $url, 0, $m2[0][1] );
1550 # Move trailing punctuation to $trail
1552 # If there is no left bracket, then consider right brackets fair game too
1553 if ( strpos( $url,
'(' ) ===
false ) {
1557 $urlRev = strrev( $url );
1558 $numSepChars = strspn( $urlRev, $sep );
1559 # Don't break a trailing HTML entity by moving the ; into $trail
1560 # This is in hot code, so use substr_compare to avoid having to
1561 # create a new string object for the comparison
1562 if ( $numSepChars && substr_compare( $url,
";", -$numSepChars, 1 ) === 0 ) {
1563 # more optimization: instead of running preg_match with a $
1564 # anchor, which can be slow, do the match on the reversed
1565 # string starting at the desired offset.
1566 # un-reversed regexp is: /&([a-z]+|#x[\da-f]+|#\d+)$/i
1567 if ( preg_match(
'/\G([a-z]+|[\da-f]+x#|\d+#)&/i', $urlRev, $m2, 0, $numSepChars ) ) {
1571 if ( $numSepChars ) {
1572 $trail = substr( $url, -$numSepChars ) . $trail;
1573 $url = substr( $url, 0, -$numSepChars );
1576 # Verify that we still have a real URL after trail removal, and
1577 # not just lone protocol
1578 if ( strlen( $trail ) >= $numPostProto ) {
1579 return $url . $trail;
1582 $url = Sanitizer::cleanUrl( $url );
1584 # Is this an external image?
1585 $text = $this->maybeMakeExternalImage( $url );
1586 if ( $text ===
false ) {
1587 # Not an image, make a link
1589 $this->getConverterLanguage()->markNoConversion( $url,
true ),
1591 $this->getExternalLinkAttribs( $url ), $this->mTitle );
1592 # Register it in the output object...
1593 $this->mOutput->addExternalLink( $url );
1595 return $text . $trail;
1607 public function doHeadings( $text ) {
1608 for ( $i = 6; $i >= 1; --$i ) {
1609 $h = str_repeat(
'=', $i );
1610 $text = preg_replace(
"/^$h(.+)$h\\s*$/m",
"<h$i>\\1</h$i>", $text );
1623 public function doAllQuotes( $text ) {
1627 $outtext .= $this->doQuotes(
$line ) .
"\n";
1629 $outtext = substr( $outtext, 0, -1 );
1640 public function doQuotes( $text ) {
1641 $arr = preg_split(
"/(''+)/", $text, -1, PREG_SPLIT_DELIM_CAPTURE );
1642 $countarr =
count( $arr );
1643 if ( $countarr == 1 ) {
1652 for ( $i = 1; $i < $countarr; $i += 2 ) {
1653 $thislen = strlen( $arr[$i] );
1657 if ( $thislen == 4 ) {
1658 $arr[$i - 1] .=
"'";
1661 } elseif ( $thislen > 5 ) {
1665 $arr[$i - 1] .= str_repeat(
"'", $thislen - 5 );
1670 if ( $thislen == 2 ) {
1672 } elseif ( $thislen == 3 ) {
1674 } elseif ( $thislen == 5 ) {
1684 if ( ( $numbold % 2 == 1 ) && ( $numitalics % 2 == 1 ) ) {
1685 $firstsingleletterword = -1;
1686 $firstmultiletterword = -1;
1688 for ( $i = 1; $i < $countarr; $i += 2 ) {
1689 if ( strlen( $arr[$i] ) == 3 ) {
1690 $x1 = substr( $arr[$i - 1], -1 );
1691 $x2 = substr( $arr[$i - 1], -2, 1 );
1692 if ( $x1 ===
' ' ) {
1693 if ( $firstspace == -1 ) {
1696 } elseif ( $x2 ===
' ' ) {
1697 $firstsingleletterword = $i;
1702 if ( $firstmultiletterword == -1 ) {
1703 $firstmultiletterword = $i;
1710 if ( $firstsingleletterword > -1 ) {
1711 $arr[$firstsingleletterword] =
"''";
1712 $arr[$firstsingleletterword - 1] .=
"'";
1713 } elseif ( $firstmultiletterword > -1 ) {
1715 $arr[$firstmultiletterword] =
"''";
1716 $arr[$firstmultiletterword - 1] .=
"'";
1717 } elseif ( $firstspace > -1 ) {
1721 $arr[$firstspace] =
"''";
1722 $arr[$firstspace - 1] .=
"'";
1731 foreach ( $arr
as $r ) {
1732 if ( ( $i % 2 ) == 0 ) {
1733 if ( $state ===
'both' ) {
1739 $thislen = strlen( $r );
1740 if ( $thislen == 2 ) {
1741 if ( $state ===
'i' ) {
1744 } elseif ( $state ===
'bi' ) {
1747 } elseif ( $state ===
'ib' ) {
1750 } elseif ( $state ===
'both' ) {
1757 } elseif ( $thislen == 3 ) {
1758 if ( $state ===
'b' ) {
1761 } elseif ( $state ===
'bi' ) {
1764 } elseif ( $state ===
'ib' ) {
1767 } elseif ( $state ===
'both' ) {
1774 } elseif ( $thislen == 5 ) {
1775 if ( $state ===
'b' ) {
1778 } elseif ( $state ===
'i' ) {
1781 } elseif ( $state ===
'bi' ) {
1784 } elseif ( $state ===
'ib' ) {
1787 } elseif ( $state ===
'both' ) {
1799 if ( $state ===
'b' || $state ===
'ib' ) {
1802 if ( $state ===
'i' || $state ===
'bi' || $state ===
'ib' ) {
1805 if ( $state ===
'bi' ) {
1809 if ( $state ===
'both' &&
$buffer ) {
1828 public function replaceExternalLinks( $text ) {
1829 $bits = preg_split( $this->mExtLinkBracketedRegex, $text, -1, PREG_SPLIT_DELIM_CAPTURE );
1830 if ( $bits ===
false ) {
1831 throw new MWException(
"PCRE needs to be compiled with "
1832 .
"--enable-unicode-properties in order for MediaWiki to function" );
1834 $s = array_shift( $bits );
1837 while ( $i <
count( $bits ) ) {
1840 $text = $bits[$i++];
1841 $trail = $bits[$i++];
1843 # The characters '<' and '>' (which were escaped by
1844 # removeHTMLtags()) should not be included in
1845 # URLs, per RFC 2396.
1847 if ( preg_match(
'/&(lt|gt);/', $url, $m2, PREG_OFFSET_CAPTURE ) ) {
1848 $text = substr( $url, $m2[0][1] ) .
' ' . $text;
1849 $url = substr( $url, 0, $m2[0][1] );
1852 # If the link text is an image URL, replace it with an <img> tag
1853 # This happened by accident in the original parser, but some people used it extensively
1854 $img = $this->maybeMakeExternalImage( $text );
1855 if ( $img !==
false ) {
1861 # Set linktype for CSS - if URL==text, link is essentially free
1862 $linktype = ( $text === $url ) ?
'free' :
'text';
1864 # No link text, e.g. [http://domain.tld/some.link]
1865 if ( $text ==
'' ) {
1867 $langObj = $this->getTargetLanguage();
1868 $text =
'[' . $langObj->formatNum( ++$this->mAutonumber ) .
']';
1869 $linktype =
'autonumber';
1871 # Have link text, e.g. [http://domain.tld/some.link text]s
1876 $text = $this->getConverterLanguage()->markNoConversion( $text );
1878 $url = Sanitizer::cleanUrl( $url );
1880 # Use the encoded URL
1881 # This means that users can paste URLs directly into the text
1882 # Funny characters like ö aren't valid in URLs anyway
1883 # This was changed in August 2004
1885 $this->getExternalLinkAttribs( $url ), $this->mTitle ) . $dtrail . $trail;
1887 # Register link in the output object.
1888 $this->mOutput->addExternalLink( $url );
1903 public static function getExternalLinkRel( $url =
false,
$title =
null ) {
1924 public function getExternalLinkAttribs( $url ) {
1926 $rel = self::getExternalLinkRel( $url, $this->mTitle );
1928 $target = $this->mOptions->getExternalLinkTarget();
1931 if ( !in_array( $target, [
'_self',
'_parent',
'_top' ] ) ) {
1935 if ( $rel !==
'' ) {
1938 $rel .=
'noreferrer noopener';
1954 public static function normalizeLinkUrl( $url ) {
1955 # First, make sure unsafe characters are encoded
1956 $url = preg_replace_callback(
'/[\x00-\x20"<>\[\\\\\]^`{|}\x7F-\xFF]/',
1958 return rawurlencode( $m[0] );
1964 $end = strlen( $url );
1966 # Fragment part - 'fragment'
1967 $start = strpos( $url,
'#' );
1968 if ( $start !==
false && $start < $end ) {
1969 $ret = self::normalizeUrlComponent(
1970 substr( $url, $start, $end - $start ),
'"#%<>[\]^`{|}' ) .
$ret;
1974 # Query part - 'query' minus &=+;
1975 $start = strpos( $url,
'?' );
1976 if ( $start !==
false && $start < $end ) {
1977 $ret = self::normalizeUrlComponent(
1978 substr( $url, $start, $end - $start ),
'"#%<>[\]^`{|}&=+;' ) .
$ret;
1982 # Scheme and path part - 'pchar'
1983 # (we assume no userinfo or encoded colons in the host)
1984 $ret = self::normalizeUrlComponent(
1985 substr( $url, 0, $end ),
'"#%<>[\]^`{|}/?' ) .
$ret;
1990 private static function normalizeUrlComponent( $component, $unsafe ) {
1991 $callback =
function (
$matches )
use ( $unsafe ) {
1993 $ord = ord( $char );
1994 if ( $ord > 32 && $ord < 127 && strpos( $unsafe, $char ) ===
false ) {
1998 # Leave it escaped, but use uppercase for a-f
2002 return preg_replace_callback(
'/%[0-9A-Fa-f]{2}/', $callback, $component );
2013 private function maybeMakeExternalImage( $url ) {
2014 $imagesfrom = $this->mOptions->getAllowExternalImagesFrom();
2015 $imagesexception = !empty( $imagesfrom );
2017 # $imagesfrom could be either a single string or an array of strings, parse out the latter
2018 if ( $imagesexception && is_array( $imagesfrom ) ) {
2019 $imagematch =
false;
2020 foreach ( $imagesfrom
as $match ) {
2021 if ( strpos( $url, $match ) === 0 ) {
2026 } elseif ( $imagesexception ) {
2027 $imagematch = ( strpos( $url, $imagesfrom ) === 0 );
2029 $imagematch =
false;
2032 if ( $this->mOptions->getAllowExternalImages()
2033 || ( $imagesexception && $imagematch )
2035 if ( preg_match( self::EXT_IMAGE_REGEX, $url ) ) {
2040 if ( !$text && $this->mOptions->getEnableImageWhitelist()
2041 && preg_match( self::EXT_IMAGE_REGEX, $url )
2043 $whitelist = explode(
2045 wfMessage(
'external_image_whitelist' )->inContentLanguage()->
text()
2048 foreach ( $whitelist
as $entry ) {
2049 # Sanitize the regex fragment, make it case-insensitive, ignore blank entries/comments
2050 if ( strpos( $entry,
'#' ) === 0 || $entry ===
'' ) {
2053 if ( preg_match(
'/' . str_replace(
'/',
'\\/', $entry ) .
'/i', $url ) ) {
2054 # Image matches a whitelist entry
2072 public function replaceInternalLinks(
$s ) {
2073 $this->mLinkHolders->merge( $this->replaceInternalLinks2(
$s ) );
2085 public function replaceInternalLinks2( &
$s ) {
2088 static $tc =
false, $e1, $e1_img;
2089 # the % is needed to support urlencoded titles as well
2092 # Match a link having the form [[namespace:link|alternate]]trail
2093 $e1 =
"/^([{$tc}]+)(?:\\|(.+?))?]](.*)\$/sD";
2094 # Match cases where there is no "]]", which might still be images
2095 $e1_img =
"/^([{$tc}]+)\\|(.*)\$/sD";
2100 # split the entire text string on occurrences of [[
2102 # get the first element (all text up to first [[), and remove the space we added
2105 $line = $a->current(); # Workaround
for broken ArrayIterator::next()
that returns "
void"
2106 $s = substr(
$s, 1 );
2108 $useLinkPrefixExtension = $this->getTargetLanguage()->linkPrefixExtension();
2110 if ( $useLinkPrefixExtension ) {
2111 # Match the end of a line for a word that's not followed by whitespace,
2112 # e.g. in the case of 'The Arab al[[Razi]]', 'al' will be matched
2115 $e2 =
"/^((?>.*[^$charset]|))(.+)$/sDu";
2118 if ( is_null( $this->mTitle ) ) {
2119 throw new MWException( __METHOD__ .
": \$this->mTitle is null\n" );
2121 $nottalk = !$this->mTitle->isTalkPage();
2123 if ( $useLinkPrefixExtension ) {
2125 if ( preg_match( $e2,
$s, $m ) ) {
2126 $first_prefix = $m[2];
2128 $first_prefix =
false;
2134 $useSubpages = $this->areSubpagesAllowed();
2137 # Loop for each link
2138 for ( ;
$line !==
false &&
$line !==
null; $a->next(),
$line = $a->current() ) {
2141 # Check for excessive memory usage
2142 if ( $holders->isBig() ) {
2144 # Do the existence check, replace the link holders and clear the array
2145 $holders->replace(
$s );
2149 if ( $useLinkPrefixExtension ) {
2150 if ( preg_match( $e2,
$s, $m ) ) {
2157 if ( $first_prefix ) {
2158 $prefix = $first_prefix;
2159 $first_prefix =
false;
2163 $might_be_img =
false;
2167 # If we get a ] at the beginning of $m[3] that means we have a link that's something like:
2168 # [[Image:Foo.jpg|[http://example.com desc]]] <- having three ] in a row fucks up,
2169 # the real problem is with the $e1 regex
2171 # Still some problems for cases where the ] is meant to be outside punctuation,
2172 # and no image is in sight. See T4095.
2174 && substr( $m[3], 0, 1 ) ===
']'
2175 && strpos( $text,
'[' ) !==
false
2177 $text .=
']'; #
so that replaceExternalLinks($text) works
later
2178 $m[3] = substr( $m[3], 1 );
2180 # fix up urlencoded title texts
2181 if ( strpos( $m[1],
'%' ) !==
false ) {
2182 # Should anchors '#' also be rejected?
2183 $m[1] = str_replace( [
'<',
'>' ], [
'<',
'>' ], rawurldecode( $m[1] ) );
2186 } elseif ( preg_match( $e1_img,
$line, $m ) ) {
2187 # Invalid, but might be an image with a link in its caption
2188 $might_be_img =
true;
2190 if ( strpos( $m[1],
'%' ) !==
false ) {
2191 $m[1] = str_replace( [
'<',
'>' ], [
'<',
'>' ], rawurldecode( $m[1] ) );
2199 $origLink = ltrim( $m[1],
' ' );
2201 # Don't allow internal links to pages containing
2202 # PROTO: where PROTO is a valid URL protocol; these
2203 # should be external links.
2204 if ( preg_match(
'/^(?i:' . $this->mUrlProtocols .
')/', $origLink ) ) {
2209 # Make subpage if necessary
2210 if ( $useSubpages ) {
2211 $link = $this->maybeDoSubpageLink( $origLink, $text );
2216 $unstrip = $this->mStripState->unstripNoWiki(
$link );
2218 if ( $nt ===
null ) {
2223 $ns = $nt->getNamespace();
2224 $iw = $nt->getInterwiki();
2226 $noforce = ( substr( $origLink, 0, 1 ) !==
':' );
2228 if ( $might_be_img ) { #
if this is actually an invalid
link
2229 if ( $ns ==
NS_FILE && $noforce ) { # but might be an image
2232 # look at the next 'line' to see if we can close it there
2234 $next_line = $a->current();
2235 if ( $next_line ===
false || $next_line ===
null ) {
2238 $m = explode(
']]', $next_line, 3 );
2239 if (
count( $m ) == 3 ) {
2240 # the first ]] closes the inner link, the second the image
2242 $text .=
"[[{$m[0]}]]{$m[1]}";
2245 } elseif (
count( $m ) == 2 ) {
2246 # if there's exactly one ]] that's fine, we'll keep looking
2247 $text .=
"[[{$m[0]}]]{$m[1]}";
2249 # if $next_line is invalid too, we need look no further
2250 $text .=
'[[' . $next_line;
2255 # we couldn't find the end of this imageLink, so output it raw
2256 # but don't ignore what might be perfectly normal links in the text we've examined
2257 $holders->merge( $this->replaceInternalLinks2( $text ) );
2258 $s .=
"{$prefix}[[$link|$text";
2259 # note: no $trail, because without an end, there *is* no trail
2262 }
else { #
it's not an image, so output it raw
2263 $s .= "{$prefix}[[$link|$text";
2264 # note: no $trail, because without an end, there *is* no trail
2269 $wasblank = ( $text == '' );
2273 # Strip off leading ':
'
2274 $text = substr( $text, 1 );
2277 # T6598 madness. Handle the quotes only if they come from the alternate part
2278 # [[Lista d''e paise d''o munno]] -> <a href="...">Lista d''e paise d''o munno</a>
2279 # [[Criticism of Harry Potter|Criticism of ''Harry Potter'']]
2280 # -> <a href="Criticism of Harry Potter">Criticism of <i>Harry Potter</i></a>
2281 $text = $this->doQuotes( $text );
2284 # Link not escaped by : , create the various objects
2285 if ( $noforce && !$nt->wasLocalInterwiki() ) {
2288 $iw && $this->mOptions->getInterwikiMagic() && $nottalk && (
2289 Language::fetchLanguageName( $iw, null, 'mw
' ) ||
2290 in_array( $iw, $wgExtraInterlanguageLinkPrefixes )
2293 # T26502: filter duplicates
2294 if ( !isset( $this->mLangLinkLanguages[$iw] ) ) {
2295 $this->mLangLinkLanguages[$iw] = true;
2296 $this->mOutput->addLanguageLink( $nt->getFullText() );
2299 $s = rtrim( $s . $prefix );
2300 $s .= trim( $trail, "\n" ) == '' ? '' : $prefix . $trail;
2304 if ( $ns == NS_FILE ) {
2305 if ( !wfIsBadImage( $nt->getDBkey(), $this->mTitle ) ) {
2307 # if no parameters were passed, $text
2308 # becomes something like "File:Foo.png",
2309 # which we don't want to pass
on to the
2313 # recursively parse links inside the image caption
2314 # actually, this will parse them in any other parameters, too,
2315 # but it might be hard to fix that, and it doesn't matter ATM
2316 $text = $this->replaceExternalLinks( $text );
2317 $holders->merge( $this->replaceInternalLinks2( $text ) );
2319 # cloak any absolute URLs inside the image markup, so replaceExternalLinks() won't touch them
2320 $s .= $prefix . $this->armorLinks(
2321 $this->makeImage( $nt, $text, $holders ) ) . $trail;
2325 $s = rtrim(
$s .
"\n" ); # T2087
2328 $sortkey = $this->getDefaultSort();
2332 $sortkey = Sanitizer::decodeCharReferences( $sortkey );
2333 $sortkey = str_replace(
"\n",
'', $sortkey );
2334 $sortkey = $this->getConverterLanguage()->convertCategoryKey( $sortkey );
2335 $this->mOutput->addCategory( $nt->getDBkey(), $sortkey );
2340 $s .= trim( $prefix . $trail,
"\n" ) ==
'' ?
'' : $prefix . $trail;
2346 # Self-link checking. For some languages, variants of the title are checked in
2347 # LinkHolderArray::doVariants() to allow batching the existence checks necessary
2348 # for linking to a different variant.
2349 if ( $ns !=
NS_SPECIAL && $nt->equals( $this->mTitle ) && !$nt->hasFragment() ) {
2354 # NS_MEDIA is a pseudo-namespace for linking directly to a file
2355 # @todo FIXME: Should do batch file existence checks, see comment below
2357 # Give extensions a chance to select the file revision for us
2361 [ $this, $nt, &
$options, &$descQuery ] );
2362 # Fetch and register the file (file title may be different via hooks)
2363 list( $file, $nt ) = $this->fetchFileAndTitle( $nt,
$options );
2364 # Cloak with NOPARSE to avoid replacement in replaceExternalLinks
2365 $s .= $prefix . $this->armorLinks(
2370 # Some titles, such as valid special pages or files in foreign repos, should
2371 # be shown as bluelinks even though they're not included in the page table
2372 # @todo FIXME: isAlwaysKnown() can be expensive for file links; we should really do
2373 # batch file existence checks for NS_FILE and NS_MEDIA
2374 if ( $iw ==
'' && $nt->isAlwaysKnown() ) {
2375 $this->mOutput->addLink( $nt );
2376 $s .= $this->makeKnownLinkHolder( $nt, $text, $trail, $prefix );
2378 # Links will be added to the output link list after checking
2379 $s .= $holders->makeHolder( $nt, $text, [], $trail, $prefix );
2398 protected function makeKnownLinkHolder( $nt, $text =
'', $trail =
'', $prefix =
'' ) {
2401 if ( $text ==
'' ) {
2402 $text = htmlspecialchars( $nt->getPrefixedText() );
2405 $link = $this->getLinkRenderer()->makeKnownLink(
2406 $nt,
new HtmlArmor(
"$prefix$text$inside" )
2409 return $this->armorLinks(
$link ) . $trail;
2422 public function armorLinks( $text ) {
2423 return preg_replace(
'/\b((?i)' . $this->mUrlProtocols .
')/',
2424 self::MARKER_PREFIX .
"NOPARSE$1", $text );
2431 public function areSubpagesAllowed() {
2432 # Some namespaces don't allow subpages
2444 public function maybeDoSubpageLink( $target, &$text ) {
2456 public function doBlockLevels( $text, $linestart ) {
2471 public function getVariableValue( $index, $frame =
false ) {
2475 if ( is_null( $this->mTitle ) ) {
2480 throw new MWException( __METHOD__ .
' Should only be '
2481 .
' called while parsing (no title set)' );
2491 if (
Hooks::run(
'ParserGetVariableValueVarCache', [ &
$parser, &$this->mVarCache ] ) ) {
2492 if ( isset( $this->mVarCache[$index] ) ) {
2493 return $this->mVarCache[$index];
2497 $ts =
wfTimestamp( TS_UNIX, $this->mOptions->getTimestamp() );
2500 $pageLang = $this->getFunctionLang();
2506 case 'currentmonth':
2509 case 'currentmonth1':
2512 case 'currentmonthname':
2515 case 'currentmonthnamegen':
2518 case 'currentmonthabbrev':
2533 case 'localmonthname':
2536 case 'localmonthnamegen':
2539 case 'localmonthabbrev':
2554 case 'fullpagename':
2557 case 'fullpagenamee':
2563 case 'subpagenamee':
2566 case 'rootpagename':
2569 case 'rootpagenamee':
2573 $this->mTitle->getRootText()
2576 case 'basepagename':
2579 case 'basepagenamee':
2583 $this->mTitle->getBaseText()
2586 case 'talkpagename':
2587 if ( $this->mTitle->canHaveTalkPage() ) {
2588 $talkPage = $this->mTitle->getTalkPage();
2594 case 'talkpagenamee':
2595 if ( $this->mTitle->canHaveTalkPage() ) {
2596 $talkPage = $this->mTitle->getTalkPage();
2602 case 'subjectpagename':
2603 $subjPage = $this->mTitle->getSubjectPage();
2606 case 'subjectpagenamee':
2607 $subjPage = $this->mTitle->getSubjectPage();
2611 $pageid = $this->getTitle()->getArticleID();
2612 if ( $pageid == 0 ) {
2613 # 0 means the page doesn't exist in the database,
2614 # which means the user is previewing a new page.
2615 # The vary-revision flag must be set, because the magic word
2616 # will have a different value once the page is saved.
2617 $this->mOutput->setFlag(
'vary-revision' );
2618 wfDebug( __METHOD__ .
": {{PAGEID}} used in a new page, setting vary-revision...\n" );
2620 $value = $pageid ? $pageid :
null;
2623 # Let the edit saving system know we should parse the page
2624 # *after* a revision ID has been assigned.
2625 $this->mOutput->setFlag(
'vary-revision-id' );
2626 wfDebug( __METHOD__ .
": {{REVISIONID}} used, setting vary-revision-id...\n" );
2627 $value = $this->mRevisionId;
2628 if ( !
$value && $this->mOptions->getSpeculativeRevIdCallback() ) {
2629 $value = call_user_func( $this->mOptions->getSpeculativeRevIdCallback() );
2630 $this->mOutput->setSpeculativeRevIdUsed(
$value );
2634 # Let the edit saving system know we should parse the page
2635 # *after* a revision ID has been assigned. This is for null edits.
2636 $this->mOutput->setFlag(
'vary-revision' );
2637 wfDebug( __METHOD__ .
": {{REVISIONDAY}} used, setting vary-revision...\n" );
2638 $value = intval( substr( $this->getRevisionTimestamp(), 6, 2 ) );
2640 case 'revisionday2':
2641 # Let the edit saving system know we should parse the page
2642 # *after* a revision ID has been assigned. This is for null edits.
2643 $this->mOutput->setFlag(
'vary-revision' );
2644 wfDebug( __METHOD__ .
": {{REVISIONDAY2}} used, setting vary-revision...\n" );
2645 $value = substr( $this->getRevisionTimestamp(), 6, 2 );
2647 case 'revisionmonth':
2648 # Let the edit saving system know we should parse the page
2649 # *after* a revision ID has been assigned. This is for null edits.
2650 $this->mOutput->setFlag(
'vary-revision' );
2651 wfDebug( __METHOD__ .
": {{REVISIONMONTH}} used, setting vary-revision...\n" );
2652 $value = substr( $this->getRevisionTimestamp(), 4, 2 );
2654 case 'revisionmonth1':
2655 # Let the edit saving system know we should parse the page
2656 # *after* a revision ID has been assigned. This is for null edits.
2657 $this->mOutput->setFlag(
'vary-revision' );
2658 wfDebug( __METHOD__ .
": {{REVISIONMONTH1}} used, setting vary-revision...\n" );
2659 $value = intval( substr( $this->getRevisionTimestamp(), 4, 2 ) );
2661 case 'revisionyear':
2662 # Let the edit saving system know we should parse the page
2663 # *after* a revision ID has been assigned. This is for null edits.
2664 $this->mOutput->setFlag(
'vary-revision' );
2665 wfDebug( __METHOD__ .
": {{REVISIONYEAR}} used, setting vary-revision...\n" );
2666 $value = substr( $this->getRevisionTimestamp(), 0, 4 );
2668 case 'revisiontimestamp':
2669 # Let the edit saving system know we should parse the page
2670 # *after* a revision ID has been assigned. This is for null edits.
2671 $this->mOutput->setFlag(
'vary-revision' );
2672 wfDebug( __METHOD__ .
": {{REVISIONTIMESTAMP}} used, setting vary-revision...\n" );
2673 $value = $this->getRevisionTimestamp();
2675 case 'revisionuser':
2676 # Let the edit saving system know we should parse the page
2677 # *after* a revision ID has been assigned for null edits.
2678 $this->mOutput->setFlag(
'vary-user' );
2679 wfDebug( __METHOD__ .
": {{REVISIONUSER}} used, setting vary-user...\n" );
2680 $value = $this->getRevisionUser();
2682 case 'revisionsize':
2683 $value = $this->getRevisionSize();
2686 $value = str_replace(
'_',
' ',
$wgContLang->getNsText( $this->mTitle->getNamespace() ) );
2691 case 'namespacenumber':
2692 $value = $this->mTitle->getNamespace();
2695 $value = $this->mTitle->canHaveTalkPage()
2696 ? str_replace(
'_',
' ', $this->mTitle->getTalkNsText() )
2700 $value = $this->mTitle->canHaveTalkPage() ?
wfUrlencode( $this->mTitle->getTalkNsText() ) :
'';
2702 case 'subjectspace':
2703 $value = str_replace(
'_',
' ', $this->mTitle->getSubjectNsText() );
2705 case 'subjectspacee':
2708 case 'currentdayname':
2721 # @bug T6594 PHP5 has it zero padded, PHP4 does not, cast to
2722 # int to remove the padding
2728 case 'localdayname':
2729 $value = $pageLang->getWeekdayName(
2737 $value = $pageLang->time(
2747 # @bug T6594 PHP5 has it zero padded, PHP4 does not, cast to
2748 # int to remove the padding
2754 case 'numberofarticles':
2757 case 'numberoffiles':
2760 case 'numberofusers':
2763 case 'numberofactiveusers':
2766 case 'numberofpages':
2769 case 'numberofadmins':
2772 case 'numberofedits':
2775 case 'currenttimestamp':
2778 case 'localtimestamp':
2781 case 'currentversion':
2796 case 'directionmark':
2797 return $pageLang->getDirMark();
2798 case 'contentlanguage':
2801 case 'pagelanguage':
2802 $value = $pageLang->getCode();
2804 case 'cascadingsources':
2810 'ParserGetVariableValueSwitch',
2811 [ &
$parser, &$this->mVarCache, &$index, &
$ret, &$frame ]
2818 $this->mVarCache[$index] =
$value;
2829 public function initialiseVariables() {
2859 public function preprocessToDom( $text,
$flags = 0 ) {
2860 $dom = $this->getPreprocessor()->preprocessToObj( $text,
$flags );
2871 public static function splitWhitespace(
$s ) {
2872 $ltrimmed = ltrim(
$s );
2873 $w1 = substr(
$s, 0, strlen(
$s ) - strlen( $ltrimmed ) );
2874 $trimmed = rtrim( $ltrimmed );
2875 $diff = strlen( $ltrimmed ) - strlen( $trimmed );
2877 $w2 = substr( $ltrimmed, -$diff );
2881 return [ $w1, $trimmed, $w2 ];
2904 public function replaceVariables( $text, $frame =
false, $argsOnly =
false ) {
2905 # Is there any text? Also, Prevent too big inclusions!
2906 $textSize = strlen( $text );
2907 if ( $textSize < 1 || $textSize > $this->mOptions->getMaxIncludeSize() ) {
2911 if ( $frame ===
false ) {
2912 $frame = $this->getPreprocessor()->newFrame();
2913 } elseif ( !( $frame instanceof
PPFrame ) ) {
2914 wfDebug( __METHOD__ .
" called using plain parameters instead of "
2915 .
"a PPFrame instance. Creating custom frame.\n" );
2916 $frame = $this->getPreprocessor()->newCustomFrame( $frame );
2919 $dom = $this->preprocessToDom( $text );
2921 $text = $frame->expand( $dom,
$flags );
2933 public static function createAssocArgs(
$args ) {
2937 $eqpos = strpos( $arg,
'=' );
2938 if ( $eqpos ===
false ) {
2939 $assocArgs[$index++] = $arg;
2941 $name = trim( substr( $arg, 0, $eqpos ) );
2942 $value = trim( substr( $arg, $eqpos + 1 ) );
2943 if (
$value ===
false ) {
2946 if (
$name !==
false ) {
2981 public function limitationWarn( $limitationType, $current =
'', $max =
'' ) {
2982 # does no harm if $current and $max are present but are unnecessary for the message
2983 # Not doing ->inLanguage( $this->mOptions->getUserLangObj() ), since this is shown
2984 # only during preview, and that would split the parser cache unnecessarily.
2985 $warning =
wfMessage(
"$limitationType-warning" )->numParams( $current, $max )
2987 $this->mOutput->addWarning( $warning );
2988 $this->addTrackingCategory(
"$limitationType-category" );
3003 public function braceSubstitution( $piece, $frame ) {
3013 $forceRawInterwiki =
false;
3015 $isChildObj =
false;
3017 $isLocalObj =
false;
3019 # Title object, where $text came from
3022 # $part1 is the bit before the first |, and must contain only title characters.
3023 # Various prefixes will be stripped from it later.
3024 $titleWithSpaces = $frame->expand( $piece[
'title'] );
3025 $part1 = trim( $titleWithSpaces );
3028 # Original title text preserved for various purposes
3029 $originalTitle = $part1;
3031 # $args is a list of argument nodes, starting from index 0, not including $part1
3032 # @todo FIXME: If piece['parts'] is null then the call to getLength()
3033 # below won't work b/c this $args isn't an object
3034 $args = (
null == $piece[
'parts'] ) ? [] : $piece[
'parts'];
3036 $profileSection =
null;
3040 $substMatch = $this->mSubstWords->matchStartAndRemove( $part1 );
3042 # Possibilities for substMatch: "subst", "safesubst" or FALSE
3043 # Decide whether to expand template or keep wikitext as-is.
3044 if ( $this->ot[
'wiki'] ) {
3045 if ( $substMatch ===
false ) {
3046 $literal =
true; # literal when
in PST with no prefix
3048 $literal =
false; # expand when
in PST with subst:
or safesubst:
3051 if ( $substMatch ==
'subst' ) {
3052 $literal =
true; # literal when
not in PST with
plain subst:
3054 $literal =
false; # expand when
not in PST with safesubst:
or no prefix
3058 $text = $frame->virtualBracketedImplode(
'{{',
'|',
'}}', $titleWithSpaces,
$args );
3065 if ( !$found &&
$args->getLength() == 0 ) {
3066 $id = $this->mVariables->matchStartToEnd( $part1 );
3067 if ( $id !==
false ) {
3068 $text = $this->getVariableValue( $id, $frame );
3076 # MSG, MSGNW and RAW
3080 if ( $mwMsgnw->matchStartAndRemove( $part1 ) ) {
3083 # Remove obsolete MSG:
3085 $mwMsg->matchStartAndRemove( $part1 );
3090 if ( $mwRaw->matchStartAndRemove( $part1 ) ) {
3091 $forceRawInterwiki =
true;
3097 $colonPos = strpos( $part1,
':' );
3098 if ( $colonPos !==
false ) {
3099 $func = substr( $part1, 0, $colonPos );
3100 $funcArgs = [ trim( substr( $part1, $colonPos + 1 ) ) ];
3101 $argsLength =
$args->getLength();
3102 for ( $i = 0; $i < $argsLength; $i++ ) {
3103 $funcArgs[] =
$args->item( $i );
3106 $result = $this->callParserFunction( $frame, $func, $funcArgs );
3107 }
catch ( Exception $ex ) {
3111 # The interface for parser functions allows for extracting
3112 # flags into the local scope. Extract any forwarded flags
3118 # Finish mangling title and then check for loops.
3119 # Set $title to a Title object and $titleText to the PDBK
3122 # Split the title into page and subpage
3124 $relative = $this->maybeDoSubpageLink( $part1, $subpage );
3125 if ( $part1 !== $relative ) {
3127 $ns = $this->mTitle->getNamespace();
3131 $titleText =
$title->getPrefixedText();
3132 # Check for language variants if the template is not found
3133 if ( $this->getConverterLanguage()->hasVariants() &&
$title->getArticleID() == 0 ) {
3134 $this->getConverterLanguage()->findVariantLink( $part1,
$title,
true );
3136 # Do recursion depth check
3137 $limit = $this->mOptions->getMaxTemplateDepth();
3138 if ( $frame->depth >= $limit ) {
3140 $text =
'<span class="error">'
3141 .
wfMessage(
'parser-template-recursion-depth-warning' )
3142 ->numParams( $limit )->inContentLanguage()->text()
3148 # Load from database
3149 if ( !$found &&
$title ) {
3150 $profileSection = $this->mProfiler->scopedProfileIn(
$title->getPrefixedDBkey() );
3151 if ( !
$title->isExternal() ) {
3152 if (
$title->isSpecialPage()
3153 && $this->mOptions->getAllowSpecialInclusion()
3154 && $this->ot[
'html']
3161 $argsLength =
$args->getLength();
3162 for ( $i = 0; $i < $argsLength; $i++ ) {
3163 $bits =
$args->item( $i )->splitArg();
3164 if ( strval( $bits[
'index'] ) ===
'' ) {
3166 $value = trim( $frame->expand( $bits[
'value'] ) );
3175 if ( $specialPage && $specialPage->maxIncludeCacheTime() === 0 ) {
3176 $context->setUser( $this->getUser() );
3181 $context->setLanguage( $this->mOptions->getUserLangObj() );
3185 $text =
$context->getOutput()->getHTML();
3186 $this->mOutput->addOutputPageMetadata(
$context->getOutput() );
3189 if ( $specialPage && $specialPage->maxIncludeCacheTime() !==
false ) {
3190 $this->mOutput->updateRuntimeAdaptiveExpiry(
3191 $specialPage->maxIncludeCacheTime()
3196 $found =
false; # access denied
3197 wfDebug( __METHOD__ .
": template inclusion denied for " .
3198 $title->getPrefixedDBkey() .
"\n" );
3201 if ( $text !==
false ) {
3207 # If the title is valid but undisplayable, make a link to it
3208 if ( !$found && ( $this->ot[
'html'] || $this->ot[
'pre'] ) ) {
3209 $text =
"[[:$titleText]]";
3212 } elseif (
$title->isTrans() ) {
3213 # Interwiki transclusion
3214 if ( $this->ot[
'html'] && !$forceRawInterwiki ) {
3215 $text = $this->interwikiTransclude(
$title,
'render' );
3218 $text = $this->interwikiTransclude(
$title,
'raw' );
3219 # Preprocess it like a template
3220 $text = $this->preprocessToDom( $text, self::PTD_FOR_INCLUSION );
3226 # Do infinite loop check
3227 # This has to be done after redirect resolution to avoid infinite loops via redirects
3228 if ( !$frame->loopCheck(
$title ) ) {
3230 $text =
'<span class="error">'
3231 .
wfMessage(
'parser-template-loop-warning', $titleText )->inContentLanguage()->text()
3233 $this->addTrackingCategory(
'template-loop-category' );
3234 $this->mOutput->addWarning(
wfMessage(
'template-loop-warning',
3236 wfDebug( __METHOD__ .
": template loop broken at '$titleText'\n" );
3240 # If we haven't found text to substitute by now, we're done
3241 # Recover the source wikitext and return it
3243 $text = $frame->virtualBracketedImplode(
'{{',
'|',
'}}', $titleWithSpaces,
$args );
3244 if ( $profileSection ) {
3245 $this->mProfiler->scopedProfileOut( $profileSection );
3247 return [
'object' => $text ];
3250 # Expand DOM-style return values in a child frame
3251 if ( $isChildObj ) {
3252 # Clean up argument array
3257 } elseif ( $titleText !==
false && $newFrame->isEmpty() ) {
3258 # Expansion is eligible for the empty-frame cache
3259 $text = $newFrame->cachedExpand( $titleText, $text );
3261 # Uncached expansion
3262 $text = $newFrame->expand( $text );
3265 if ( $isLocalObj && $nowiki ) {
3267 $isLocalObj =
false;
3270 if ( $profileSection ) {
3271 $this->mProfiler->scopedProfileOut( $profileSection );
3274 # Replace raw HTML by a placeholder
3276 $text = $this->insertStripItem( $text );
3277 } elseif ( $nowiki && ( $this->ot[
'html'] || $this->ot[
'pre'] ) ) {
3278 # Escape nowiki-style return values
3280 } elseif ( is_string( $text )
3281 && !$piece[
'lineStart']
3282 && preg_match(
'/^(?:{\\||:|;|#|\*)/', $text )
3284 # T2529: if the template begins with a table or block-level
3285 # element, it should be treated as beginning a new line.
3286 # This behavior is somewhat controversial.
3287 $text =
"\n" . $text;
3290 if ( is_string( $text ) && !$this->incrementIncludeSize(
'post-expand', strlen( $text ) ) ) {
3291 # Error, oversize inclusion
3292 if ( $titleText !==
false ) {
3293 # Make a working, properly escaped link if possible (T25588)
3294 $text =
"[[:$titleText]]";
3296 # This will probably not be a working link, but at least it may
3297 # provide some hint of where the problem is
3298 preg_replace(
'/^:/',
'', $originalTitle );
3299 $text =
"[[:$originalTitle]]";
3301 $text .= $this->insertStripItem(
'<!-- WARNING: template omitted, '
3302 .
'post-expand include size too large -->' );
3303 $this->limitationWarn(
'post-expand-template-inclusion' );
3306 if ( $isLocalObj ) {
3307 $ret = [
'object' => $text ];
3309 $ret = [
'text' => $text ];
3334 public function callParserFunction( $frame, $function,
array $args = [] ) {
3337 # Case sensitive functions
3338 if ( isset( $this->mFunctionSynonyms[1][$function] ) ) {
3339 $function = $this->mFunctionSynonyms[1][$function];
3341 # Case insensitive functions
3343 if ( isset( $this->mFunctionSynonyms[0][$function] ) ) {
3344 $function = $this->mFunctionSynonyms[0][$function];
3346 return [
'found' =>
false ];
3350 list( $callback,
$flags ) = $this->mFunctionHooks[$function];
3357 # Convert arguments to PPNodes and collect for appending to $allArgs
3359 foreach (
$args as $k => $v ) {
3360 if ( $v instanceof
PPNode || $k === 0 ) {
3363 $funcArgs[] = $this->mPreprocessor->newPartNodeArray( [ $k => $v ] )->item( 0 );
3367 # Add a frame parameter, and pass the arguments as an array
3368 $allArgs[] = $frame;
3369 $allArgs[] = $funcArgs;
3371 # Convert arguments to plain text and append to $allArgs
3372 foreach (
$args as $k => $v ) {
3373 if ( $v instanceof
PPNode ) {
3374 $allArgs[] = trim( $frame->expand( $v ) );
3375 } elseif ( is_int( $k ) && $k >= 0 ) {
3376 $allArgs[] = trim( $v );
3378 $allArgs[] = trim(
"$k=$v" );
3383 $result = call_user_func_array( $callback, $allArgs );
3385 # The interface for function hooks allows them to return a wikitext
3386 # string or an array containing the string and any flags. This mungs
3387 # things around to match what this method should return.
3404 $preprocessFlags = 0;
3405 if ( isset(
$result[
'noparse'] ) ) {
3406 $noparse =
$result[
'noparse'];
3408 if ( isset(
$result[
'preprocessFlags'] ) ) {
3409 $preprocessFlags =
$result[
'preprocessFlags'];
3413 $result[
'text'] = $this->preprocessToDom(
$result[
'text'], $preprocessFlags );
3428 public function getTemplateDom(
$title ) {
3430 $titleText =
$title->getPrefixedDBkey();
3432 if ( isset( $this->mTplRedirCache[$titleText] ) ) {
3433 list( $ns, $dbk ) = $this->mTplRedirCache[$titleText];
3435 $titleText =
$title->getPrefixedDBkey();
3437 if ( isset( $this->mTplDomCache[$titleText] ) ) {
3438 return [ $this->mTplDomCache[$titleText],
$title ];
3441 # Cache miss, go to the database
3444 if ( $text ===
false ) {
3445 $this->mTplDomCache[$titleText] =
false;
3446 return [
false,
$title ];
3449 $dom = $this->preprocessToDom( $text, self::PTD_FOR_INCLUSION );
3450 $this->mTplDomCache[$titleText] = $dom;
3452 if ( !
$title->equals( $cacheTitle ) ) {
3453 $this->mTplRedirCache[$cacheTitle->getPrefixedDBkey()] =
3471 public function fetchCurrentRevisionOfTitle(
$title ) {
3472 $cacheKey =
$title->getPrefixedDBkey();
3473 if ( !$this->currentRevisionCache ) {
3474 $this->currentRevisionCache =
new MapCacheLRU( 100 );
3476 if ( !$this->currentRevisionCache->has( $cacheKey ) ) {
3477 $this->currentRevisionCache->set( $cacheKey,
3479 call_user_func( $this->mOptions->getCurrentRevisionCallback(),
$title, $this )
3482 return $this->currentRevisionCache->get( $cacheKey );
3495 $pageId =
$title->getArticleID();
3496 $revId =
$title->getLatestRevID();
3511 public function fetchTemplateAndTitle(
$title ) {
3513 $templateCb = $this->mOptions->getTemplateCallback();
3514 $stuff = call_user_func( $templateCb,
$title, $this );
3516 $text = $stuff[
'text'];
3517 if ( is_string( $stuff[
'text'] ) ) {
3518 $text = strtr( $text,
"\x7f",
"?" );
3520 $finalTitle = isset( $stuff[
'finalTitle'] ) ? $stuff[
'finalTitle'] :
$title;
3521 if ( isset( $stuff[
'deps'] ) ) {
3522 foreach ( $stuff[
'deps']
as $dep ) {
3523 $this->mOutput->addTemplate( $dep[
'title'], $dep[
'page_id'], $dep[
'rev_id'] );
3524 if ( $dep[
'title']->equals( $this->getTitle() ) ) {
3527 $this->mOutput->setFlag(
'vary-revision' );
3531 return [ $text, $finalTitle ];
3539 public function fetchTemplate(
$title ) {
3540 return $this->fetchTemplateAndTitle(
$title )[0];
3552 public static function statelessFetchTemplate(
$title,
$parser =
false ) {
3553 $text = $skip =
false;
3557 # Loop to fetch the article, with up to 1 redirect
3559 for ( $i = 0; $i < 2 && is_object(
$title ); $i++ ) {
3561 # Give extensions a chance to select the revision instead
3562 $id =
false; # Assume current
3563 Hooks::run(
'BeforeParserFetchTemplateAndtitle',
3570 'page_id' =>
$title->getArticleID(),
3583 $rev_id =
$rev ?
$rev->getId() : 0;
3584 # If there is no current revision, there is no page
3585 if ( $id ===
false && !
$rev ) {
3587 $linkCache->addBadLinkObj(
$title );
3592 'page_id' =>
$title->getArticleID(),
3593 'rev_id' => $rev_id ];
3595 # We fetched a rev from a different title; register it too...
3597 'title' =>
$rev->getTitle(),
3598 'page_id' =>
$rev->getPage(),
3599 'rev_id' => $rev_id ];
3603 $content =
$rev->getContent();
3604 $text = $content ? $content->getWikitextForTransclusion() :
null;
3609 if ( $text ===
false || $text ===
null ) {
3616 if ( !$message->exists() ) {
3620 $content = $message->content();
3621 $text = $message->plain();
3630 $title = $content->getRedirectTarget();
3634 'finalTitle' => $finalTitle,
3659 $time = $file ? $file->getTimestamp() :
false;
3660 $sha1 = $file ? $file->getSha1() :
false;
3661 # Register the file as a dependency...
3662 $this->mOutput->addImage(
$title->getDBkey(),
$time, $sha1 );
3663 if ( $file && !
$title->equals( $file->getTitle() ) ) {
3664 # Update fetched file title
3665 $title = $file->getTitle();
3666 $this->mOutput->addImage(
$title->getDBkey(),
$time, $sha1 );
3668 return [ $file,
$title ];
3681 protected function fetchFileNoRegister(
$title,
$options = [] ) {
3682 if ( isset(
$options[
'broken'] ) ) {
3684 } elseif ( isset(
$options[
'sha1'] ) ) {
3700 public function interwikiTransclude(
$title, $action ) {
3704 return wfMessage(
'scarytranscludedisabled' )->inContentLanguage()->text();
3707 $url =
$title->getFullURL( [
'action' => $action ] );
3709 if ( strlen( $url ) > 255 ) {
3710 return wfMessage(
'scarytranscludetoolong' )->inContentLanguage()->text();
3712 return $this->fetchScaryTemplateMaybeFromCache( $url );
3719 public function fetchScaryTemplateMaybeFromCache( $url ) {
3723 $obj =
$dbr->selectRow(
'transcache', [
'tc_time',
'tc_contents' ],
3724 [
'tc_url' => $url,
"tc_time >= " .
$dbr->addQuotes( $tsCond ) ] );
3726 return $obj->tc_contents;
3732 $text =
$req->getContent();
3733 } elseif (
$req->getStatus() != 200 ) {
3735 return wfMessage(
'scarytranscludefailed-httpstatus' )
3736 ->params( $url,
$req->getStatus() )->inContentLanguage()->text();
3738 return wfMessage(
'scarytranscludefailed', $url )->inContentLanguage()->text();
3742 $dbw->replace(
'transcache', [
'tc_url' ], [
3744 'tc_time' => $dbw->timestamp( time() ),
3745 'tc_contents' => $text
3759 public function argSubstitution( $piece, $frame ) {
3761 $parts = $piece[
'parts'];
3762 $nameWithSpaces = $frame->expand( $piece[
'title'] );
3763 $argName = trim( $nameWithSpaces );
3765 $text = $frame->getArgument( $argName );
3766 if ( $text ===
false && $parts->getLength() > 0
3767 && ( $this->ot[
'html']
3769 || ( $this->ot[
'wiki'] && $frame->isTemplate() )
3772 # No match in frame, use the supplied default
3773 $object = $parts->item( 0 )->getChildren();
3775 if ( !$this->incrementIncludeSize(
'arg', strlen( $text ) ) ) {
3776 $error =
'<!-- WARNING: argument omitted, expansion size too large -->';
3777 $this->limitationWarn(
'post-expand-template-argument' );
3780 if ( $text ===
false && $object ===
false ) {
3782 $object = $frame->virtualBracketedImplode(
'{{{',
'|',
'}}}', $nameWithSpaces, $parts );
3784 if ( $error !==
false ) {
3787 if ( $object !==
false ) {
3788 $ret = [
'object' => $object ];
3790 $ret = [
'text' => $text ];
3811 public function extensionSubstitution(
$params, $frame ) {
3812 static $errorStr =
'<span class="error">';
3813 static $errorLen = 20;
3816 if ( substr(
$name, 0, $errorLen ) === $errorStr ) {
3822 $attrText = !isset(
$params[
'attr'] ) ? null : $frame->expand(
$params[
'attr'] );
3823 if ( substr( $attrText, 0, $errorLen ) === $errorStr ) {
3831 $content = !isset(
$params[
'inner'] ) ? null : $frame->expand(
$params[
'inner'] );
3833 $marker = self::MARKER_PREFIX .
"-$name-"
3834 . sprintf(
'%08X', $this->mMarkerIndex++ ) . self::MARKER_SUFFIX;
3836 $isFunctionTag = isset( $this->mFunctionTagHooks[strtolower(
$name )] ) &&
3837 ( $this->ot[
'html'] || $this->ot[
'pre'] );
3838 if ( $isFunctionTag ) {
3839 $markerType =
'none';
3841 $markerType =
'general';
3843 if ( $this->ot[
'html'] || $isFunctionTag ) {
3845 $attributes = Sanitizer::decodeTagAttributes( $attrText );
3846 if ( isset(
$params[
'attributes'] ) ) {
3847 $attributes = $attributes +
$params[
'attributes'];
3850 if ( isset( $this->mTagHooks[
$name] ) ) {
3851 $output = call_user_func_array( $this->mTagHooks[
$name],
3852 [ $content, $attributes, $this, $frame ] );
3853 } elseif ( isset( $this->mFunctionTagHooks[
$name] ) ) {
3854 list( $callback, ) = $this->mFunctionTagHooks[
$name];
3858 $output = call_user_func_array( $callback, [ &
$parser, $frame, $content, $attributes ] );
3860 $output =
'<span class="error">Invalid tag extension name: ' .
3861 htmlspecialchars(
$name ) .
'</span>';
3865 # Extract flags to local scope (to override $markerType)
3872 if ( is_null( $attrText ) ) {
3875 if ( isset(
$params[
'attributes'] ) ) {
3876 foreach (
$params[
'attributes']
as $attrName => $attrValue ) {
3877 $attrText .=
' ' . htmlspecialchars( $attrName ) .
'="' .
3878 htmlspecialchars( $attrValue ) .
'"';
3881 if ( $content ===
null ) {
3882 $output =
"<$name$attrText/>";
3884 $close = is_null(
$params[
'close'] ) ?
'' : $frame->expand(
$params[
'close'] );
3885 if ( substr( $close, 0, $errorLen ) === $errorStr ) {
3889 $output =
"<$name$attrText>$content$close";
3893 if ( $markerType ===
'none' ) {
3895 } elseif ( $markerType ===
'nowiki' ) {
3896 $this->mStripState->addNoWiki( $marker,
$output );
3897 } elseif ( $markerType ===
'general' ) {
3898 $this->mStripState->addGeneral( $marker,
$output );
3900 throw new MWException( __METHOD__ .
': invalid marker type' );
3912 public function incrementIncludeSize(
$type, $size ) {
3913 if ( $this->mIncludeSizes[
$type] + $size > $this->mOptions->getMaxIncludeSize() ) {
3916 $this->mIncludeSizes[
$type] += $size;
3926 public function incrementExpensiveFunctionCount() {
3927 $this->mExpensiveFunctionCount++;
3928 return $this->mExpensiveFunctionCount <= $this->mOptions->getExpensiveParserFunctionLimit();
3939 public function doDoubleUnderscore( $text ) {
3940 # The position of __TOC__ needs to be recorded
3942 if ( $mw->match( $text ) ) {
3943 $this->mShowToc =
true;
3944 $this->mForceTocPosition =
true;
3946 # Set a placeholder. At the end we'll fill it in with the TOC.
3947 $text = $mw->replace(
'<!--MWTOC-->', $text, 1 );
3949 # Only keep the first one.
3950 $text = $mw->replace(
'', $text );
3953 # Now match and remove the rest of them
3955 $this->mDoubleUnderscores = $mwa->matchAndRemove( $text );
3957 if ( isset( $this->mDoubleUnderscores[
'nogallery'] ) ) {
3958 $this->mOutput->mNoGallery =
true;
3960 if ( isset( $this->mDoubleUnderscores[
'notoc'] ) && !$this->mForceTocPosition ) {
3961 $this->mShowToc =
false;
3963 if ( isset( $this->mDoubleUnderscores[
'hiddencat'] )
3966 $this->addTrackingCategory(
'hidden-category-category' );
3968 # (T10068) Allow control over whether robots index a page.
3969 # __INDEX__ always overrides __NOINDEX__, see T16899
3970 if ( isset( $this->mDoubleUnderscores[
'noindex'] ) && $this->mTitle->canUseNoindex() ) {
3971 $this->mOutput->setIndexPolicy(
'noindex' );
3972 $this->addTrackingCategory(
'noindex-category' );
3974 if ( isset( $this->mDoubleUnderscores[
'index'] ) && $this->mTitle->canUseNoindex() ) {
3975 $this->mOutput->setIndexPolicy(
'index' );
3976 $this->addTrackingCategory(
'index-category' );
3979 # Cache all double underscores in the database
3980 foreach ( $this->mDoubleUnderscores
as $key => $val ) {
3981 $this->mOutput->setProperty( $key,
'' );
3992 public function addTrackingCategory( $msg ) {
3993 return $this->mOutput->addTrackingCategory( $msg, $this->mTitle );
4012 public function formatHeadings( $text, $origText, $isMain =
true ) {
4015 # Inhibit editsection links if requested in the page
4016 if ( isset( $this->mDoubleUnderscores[
'noeditsection'] ) ) {
4017 $maybeShowEditLink = $showEditLink =
false;
4019 $maybeShowEditLink =
true;
4020 $showEditLink = $this->mOptions->getEditSection();
4022 if ( $showEditLink ) {
4023 $this->mOutput->setEditSectionTokens(
true );
4026 # Get all headlines for numbering them and adding funky stuff like [edit]
4027 # links - this is for later, but we need the number of headlines right now
4029 $numMatches = preg_match_all(
4030 '/<H(?P<level>[1-6])(?P<attrib>.*?>)\s*(?P<header>[\s\S]*?)\s*<\/H[1-6] *>/i',
4035 # if there are fewer than 4 headlines in the article, do not show TOC
4036 # unless it's been explicitly enabled.
4037 $enoughToc = $this->mShowToc &&
4038 ( ( $numMatches >= 4 ) || $this->mForceTocPosition );
4040 # Allow user to stipulate that a page should have a "new section"
4041 # link added via __NEWSECTIONLINK__
4042 if ( isset( $this->mDoubleUnderscores[
'newsectionlink'] ) ) {
4043 $this->mOutput->setNewSection(
true );
4046 # Allow user to remove the "new section"
4047 # link via __NONEWSECTIONLINK__
4048 if ( isset( $this->mDoubleUnderscores[
'nonewsectionlink'] ) ) {
4049 $this->mOutput->hideNewSection(
true );
4052 # if the string __FORCETOC__ (not case-sensitive) occurs in the HTML,
4053 # override above conditions and always show TOC above first header
4054 if ( isset( $this->mDoubleUnderscores[
'forcetoc'] ) ) {
4055 $this->mShowToc =
true;
4063 # Ugh .. the TOC should have neat indentation levels which can be
4064 # passed to the skin functions. These are determined here
4068 $sublevelCount = [];
4074 $markerRegex = self::MARKER_PREFIX .
"-h-(\d+)-" . self::MARKER_SUFFIX;
4075 $baseTitleText = $this->mTitle->getPrefixedDBkey();
4076 $oldType = $this->mOutputType;
4078 $frame = $this->getPreprocessor()->newFrame();
4079 $root = $this->preprocessToDom( $origText );
4080 $node = $root->getFirstChild();
4085 $headlines = $numMatches !==
false ?
$matches[3] : [];
4087 foreach ( $headlines
as $headline ) {
4088 $isTemplate =
false;
4090 $sectionIndex =
false;
4092 $markerMatches = [];
4093 if ( preg_match(
"/^$markerRegex/", $headline, $markerMatches ) ) {
4094 $serial = $markerMatches[1];
4095 list( $titleText, $sectionIndex ) = $this->mHeadings[$serial];
4096 $isTemplate = ( $titleText != $baseTitleText );
4097 $headline = preg_replace(
"/^$markerRegex\\s*/",
"", $headline );
4101 $prevlevel = $level;
4103 $level =
$matches[1][$headlineCount];
4105 if ( $level > $prevlevel ) {
4106 # Increase TOC level
4108 $sublevelCount[$toclevel] = 0;
4110 $prevtoclevel = $toclevel;
4114 } elseif ( $level < $prevlevel && $toclevel > 1 ) {
4115 # Decrease TOC level, find level to jump to
4117 for ( $i = $toclevel; $i > 0; $i-- ) {
4118 if ( $levelCount[$i] == $level ) {
4119 # Found last matching level
4122 } elseif ( $levelCount[$i] < $level ) {
4123 # Found first matching level below current level
4133 # Unindent only if the previous toc level was shown :p
4135 $prevtoclevel = $toclevel;
4141 # No change in level, end TOC line
4147 $levelCount[$toclevel] = $level;
4149 # count number of headlines for each level
4150 $sublevelCount[$toclevel]++;
4152 for ( $i = 1; $i <= $toclevel; $i++ ) {
4153 if ( !empty( $sublevelCount[$i] ) ) {
4157 $numbering .= $this->getTargetLanguage()->formatNum( $sublevelCount[$i] );
4162 # The safe header is a version of the header text safe to use for links
4164 # Remove link placeholders by the link text.
4165 # <!--LINK number-->
4167 # link text with suffix
4168 # Do this before unstrip since link text can contain strip markers
4169 $safeHeadline = $this->replaceLinkHoldersText( $headline );
4171 # Avoid insertion of weird stuff like <math> by expanding the relevant sections
4172 $safeHeadline = $this->mStripState->unstripBoth( $safeHeadline );
4174 # Strip out HTML (first regex removes any tag not allowed)
4176 # * <sup> and <sub> (T10393)
4180 # * <span dir="rtl"> and <span dir="ltr"> (T37167)
4181 # * <s> and <strike> (T35715)
4182 # We strip any parameter from accepted tags (second regex), except dir="rtl|ltr" from <span>,
4183 # to allow setting directionality in toc items.
4184 $tocline = preg_replace(
4186 '#<(?!/?(span|sup|sub|bdi|i|b|s|strike)(?: [^>]*)?>).*?>#',
4187 '#<(/?(?:span(?: dir="(?:rtl|ltr)")?|sup|sub|bdi|i|b|s|strike))(?: .*?)?>#'
4193 # Strip '<span></span>', which is the result from the above if
4194 # <span id="foo"></span> is used to produce an additional anchor
4196 $tocline = str_replace(
'<span></span>',
'', $tocline );
4198 $tocline = trim( $tocline );
4200 # For the anchor, strip out HTML-y stuff period
4201 $safeHeadline = preg_replace(
'/<.*?>/',
'', $safeHeadline );
4202 $safeHeadline = Sanitizer::normalizeSectionNameWhitespace( $safeHeadline );
4204 # Save headline for section edit hint before it's escaped
4205 $headlineHint = $safeHeadline;
4207 # Decode HTML entities
4208 $safeHeadline = Sanitizer::decodeCharReferences( $safeHeadline );
4209 $fallbackHeadline = Sanitizer::escapeIdForAttribute( $safeHeadline, Sanitizer::ID_FALLBACK );
4210 $linkAnchor = Sanitizer::escapeIdForLink( $safeHeadline );
4211 $safeHeadline = Sanitizer::escapeIdForAttribute( $safeHeadline, Sanitizer::ID_PRIMARY );
4212 if ( $fallbackHeadline === $safeHeadline ) {
4213 # No reason to have both (in fact, we can't)
4214 $fallbackHeadline =
false;
4217 # HTML IDs must be case-insensitively unique for IE compatibility (T12721).
4218 # @todo FIXME: We may be changing them depending on the current locale.
4219 $arrayKey = strtolower( $safeHeadline );
4220 if ( $fallbackHeadline ===
false ) {
4221 $fallbackArrayKey =
false;
4223 $fallbackArrayKey = strtolower( $fallbackHeadline );
4226 # Create the anchor for linking from the TOC to the section
4227 $anchor = $safeHeadline;
4228 $fallbackAnchor = $fallbackHeadline;
4229 if ( isset( $refers[$arrayKey] ) ) {
4231 for ( $i = 2; isset( $refers[
"${arrayKey}_$i"] ); ++$i );
4234 $linkAnchor .=
"_$i";
4235 $refers[
"${arrayKey}_$i"] =
true;
4237 $refers[$arrayKey] =
true;
4239 if ( $fallbackHeadline !==
false && isset( $refers[$fallbackArrayKey] ) ) {
4241 for ( $i = 2; isset( $refers[
"${fallbackArrayKey}_$i"] ); ++$i );
4243 $fallbackAnchor .=
"_$i";
4244 $refers[
"${fallbackArrayKey}_$i"] =
true;
4246 $refers[$fallbackArrayKey] =
true;
4249 # Don't number the heading if it is the only one (looks silly)
4250 if (
count(
$matches[3] ) > 1 && $this->mOptions->getNumberHeadings() ) {
4251 # the two are different if the line contains a link
4254 [
'class' =>
'mw-headline-number' ],
4256 ) .
' ' . $headline;
4261 $numbering, $toclevel, ( $isTemplate ?
false : $sectionIndex ) );
4264 # Add the section to the section tree
4265 # Find the DOM node for this header
4266 $noOffset = ( $isTemplate || $sectionIndex ===
false );
4267 while ( $node && !$noOffset ) {
4268 if ( $node->getName() ===
'h' ) {
4269 $bits = $node->splitHeading();
4270 if ( $bits[
'i'] == $sectionIndex ) {
4274 $byteOffset += mb_strlen( $this->mStripState->unstripBoth(
4276 $node = $node->getNextSibling();
4279 'toclevel' => $toclevel,
4282 'number' => $numbering,
4283 'index' => ( $isTemplate ?
'T-' :
'' ) . $sectionIndex,
4284 'fromtitle' => $titleText,
4285 'byteoffset' => ( $noOffset ?
null : $byteOffset ),
4286 'anchor' => $anchor,
4289 # give headline the correct <h#> tag
4290 if ( $maybeShowEditLink && $sectionIndex !==
false ) {
4292 if ( $isTemplate ) {
4293 # Put a T flag in the section identifier, to indicate to extractSections()
4294 # that sections inside <includeonly> should be counted.
4295 $editsectionPage = $titleText;
4296 $editsectionSection =
"T-$sectionIndex";
4297 $editsectionContent =
null;
4299 $editsectionPage = $this->mTitle->getPrefixedText();
4300 $editsectionSection = $sectionIndex;
4301 $editsectionContent = $headlineHint;
4315 $editlink =
'<mw:editsection page="' . htmlspecialchars( $editsectionPage );
4316 $editlink .=
'" section="' . htmlspecialchars( $editsectionSection ) .
'"';
4317 if ( $editsectionContent !==
null ) {
4318 $editlink .=
'>' . $editsectionContent .
'</mw:editsection>';
4326 $matches[
'attrib'][$headlineCount], $anchor, $headline,
4327 $editlink, $fallbackAnchor );
4332 $this->setOutputType( $oldType );
4334 # Never ever show TOC if no headers
4335 if ( $numVisible < 1 ) {
4344 $this->mOutput->setTOCHTML( $toc );
4345 $toc = self::TOC_START . $toc . self::TOC_END;
4349 $this->mOutput->setSections( $tocraw );
4352 # split up and insert constructed headlines
4353 $blocks = preg_split(
'/<H[1-6].*?>[\s\S]*?<\/H[1-6]>/i', $text );
4358 foreach ( $blocks
as $block ) {
4360 if ( empty( $head[$i - 1] ) ) {
4361 $sections[$i] = $block;
4363 $sections[$i] = $head[$i - 1] . $block;
4376 Hooks::run(
'ParserSectionCreate', [ $this, $i, &$sections[$i], $showEditLink ] );
4381 if ( $enoughToc && $isMain && !$this->mForceTocPosition ) {
4384 $sections[0] = $sections[0] . $toc .
"\n";
4387 $full .= implode(
'', $sections );
4389 if ( $this->mForceTocPosition ) {
4390 return str_replace(
'<!--MWTOC-->', $toc, $full );
4410 if ( $clearState ) {
4411 $magicScopeVariable = $this->lock();
4414 $this->setUser(
$user );
4417 $text = str_replace(
"\000",
'', $text );
4424 if (
$options->getPreSaveTransform() ) {
4425 $text = $this->pstPass2( $text,
$user );
4427 $text = $this->mStripState->unstripBoth( $text );
4429 $this->setUser(
null ); # Reset
4442 private function pstPass2( $text,
$user ) {
4445 # Note: This is the timestamp saved as hardcoded wikitext to
4446 # the database, we use $wgContLang here in order to give
4447 # everyone the same signature and use the default one rather
4448 # than the one selected in each user's preferences.
4450 $ts = $this->mOptions->getTimestamp();
4452 $ts = $timestamp->format(
'YmdHis' );
4453 $tzMsg = $timestamp->getTimezoneMessage()->inContentLanguage()->text();
4455 $d =
$wgContLang->timeanddate( $ts,
false,
false ) .
" ($tzMsg)";
4457 # Variable replacement
4458 # Because mOutputType is OT_WIKI, this will only process {{subst:xxx}} type tags
4459 $text = $this->replaceVariables( $text );
4461 # This works almost by chance, as the replaceVariables are done before the getUserSig(),
4462 # which may corrupt this parser instance via its wfMessage()->text() call-
4465 if ( strpos( $text,
'~~~' ) !==
false ) {
4466 $sigText = $this->getUserSig(
$user );
4467 $text = strtr( $text, [
4469 '~~~~' =>
"$sigText $d",
4472 # The main two signature forms used above are time-sensitive
4473 $this->mOutput->setFlag(
'user-signature' );
4476 # Context links ("pipe tricks"): [[|name]] and [[name (context)|]]
4478 $nc =
'[ _0-9A-Za-z\x80-\xff-]'; # Namespaces can
use non-ascii!
4481 $p1 =
"/\[\[(:?$nc+:|:|)($tc+?)( ?\\($tc+\\))\\|]]/";
4483 $p4 =
"/\[\[(:?$nc+:|:|)($tc+?)( ?($tc+))\\|]]/";
4485 $p3 =
"/\[\[(:?$nc+:|:|)($tc+?)( ?\\($tc+\\)|)((?:, |,)$tc+|)\\|]]/";
4487 $p2 =
"/\[\[\\|($tc+)]]/";
4489 # try $p1 first, to turn "[[A, B (C)|]]" into "[[A, B (C)|A, B]]"
4490 $text = preg_replace( $p1,
'[[\\1\\2\\3|\\2]]', $text );
4491 $text = preg_replace( $p4,
'[[\\1\\2\\3|\\2]]', $text );
4492 $text = preg_replace( $p3,
'[[\\1\\2\\3\\4|\\2]]', $text );
4494 $t = $this->mTitle->getText();
4496 if ( preg_match(
"/^($nc+:|)$tc+?( \\($tc+\\))$/",
$t, $m ) ) {
4497 $text = preg_replace( $p2,
"[[$m[1]\\1$m[2]|\\1]]", $text );
4498 } elseif ( preg_match(
"/^($nc+:|)$tc+?(, $tc+|)$/",
$t, $m ) &&
"$m[1]$m[2]" !=
'' ) {
4499 $text = preg_replace( $p2,
"[[$m[1]\\1$m[2]|\\1]]", $text );
4501 # if there's no context, don't bother duplicating the title
4502 $text = preg_replace( $p2,
'[[\\1]]', $text );
4522 public function getUserSig( &
$user, $nickname =
false, $fancySig =
null ) {
4527 # If not given, retrieve from the user object.
4528 if ( $nickname ===
false ) {
4529 $nickname =
$user->getOption(
'nickname' );
4532 if ( is_null( $fancySig ) ) {
4533 $fancySig =
$user->getBoolOption(
'fancysig' );
4536 $nickname = $nickname ==
null ?
$username : $nickname;
4540 wfDebug( __METHOD__ .
": $username has overlong signature.\n" );
4541 } elseif ( $fancySig !==
false ) {
4542 # Sig. might contain markup; validate this
4543 if ( $this->validateSig( $nickname ) !==
false ) {
4544 # Validated; clean up (if needed) and return it
4545 return $this->cleanSig( $nickname,
true );
4547 # Failed to validate; fall back to the default
4549 wfDebug( __METHOD__ .
": $username has bad XML tags in signature.\n" );
4553 # Make sure nickname doesnt get a sig in a sig
4554 $nickname = self::cleanSigInSig( $nickname );
4556 # If we're still here, make it a link to the user page
4559 $msgName =
$user->isAnon() ?
'signature-anon' :
'signature';
4561 return wfMessage( $msgName, $userText, $nickText )->inContentLanguage()
4562 ->title( $this->getTitle() )->text();
4571 public function validateSig( $text ) {
4585 public function cleanSig( $text, $parsing =
false ) {
4588 $magicScopeVariable = $this->lock();
4592 # Option to disable this feature
4593 if ( !$this->mOptions->getCleanSignatures() ) {
4597 # @todo FIXME: Regex doesn't respect extension tags or nowiki
4598 # => Move this logic to braceSubstitution()
4600 $substRegex =
'/\{\{(?!(?:' . $substWord->getBaseRegex() .
'))/x' . $substWord->getRegexCase();
4601 $substText =
'{{' . $substWord->getSynonym( 0 );
4603 $text = preg_replace( $substRegex, $substText, $text );
4604 $text = self::cleanSigInSig( $text );
4605 $dom = $this->preprocessToDom( $text );
4606 $frame = $this->getPreprocessor()->newFrame();
4607 $text = $frame->expand( $dom );
4610 $text = $this->mStripState->unstripBoth( $text );
4622 public static function cleanSigInSig( $text ) {
4623 $text = preg_replace(
'/~{3,5}/',
'', $text );
4637 $outputType, $clearState =
true
4649 $outputType, $clearState =
true
4651 $this->setTitle(
$title );
4653 $this->setOutputType( $outputType );
4654 if ( $clearState ) {
4655 $this->clearState();
4668 static $executing =
false;
4670 # Guard against infinite recursion
4711 public function setHook( $tag, callable $callback ) {
4712 $tag = strtolower( $tag );
4713 if ( preg_match(
'/[<>\r\n]/', $tag, $m ) ) {
4714 throw new MWException(
"Invalid character {$m[0]} in setHook('$tag', ...) call" );
4716 $oldVal = isset( $this->mTagHooks[$tag] ) ? $this->mTagHooks[$tag] :
null;
4717 $this->mTagHooks[$tag] = $callback;
4718 if ( !in_array( $tag, $this->mStripList ) ) {
4719 $this->mStripList[] = $tag;
4742 public function setTransparentTagHook( $tag, callable $callback ) {
4743 $tag = strtolower( $tag );
4744 if ( preg_match(
'/[<>\r\n]/', $tag, $m ) ) {
4745 throw new MWException(
"Invalid character {$m[0]} in setTransparentHook('$tag', ...) call" );
4747 $oldVal = isset( $this->mTransparentTagHooks[$tag] ) ? $this->mTransparentTagHooks[$tag] :
null;
4748 $this->mTransparentTagHooks[$tag] = $callback;
4756 public function clearTagHooks() {
4757 $this->mTagHooks = [];
4758 $this->mFunctionTagHooks = [];
4759 $this->mStripList = $this->mDefaultStripList;
4805 public function setFunctionHook( $id, callable $callback,
$flags = 0 ) {
4808 $oldVal = isset( $this->mFunctionHooks[$id] ) ? $this->mFunctionHooks[$id][0] :
null;
4809 $this->mFunctionHooks[$id] = [ $callback,
$flags ];
4811 # Add to function cache
4814 throw new MWException( __METHOD__ .
'() expecting a magic word identifier.' );
4817 $synonyms = $mw->getSynonyms();
4818 $sensitive = intval( $mw->isCaseSensitive() );
4820 foreach ( $synonyms
as $syn ) {
4822 if ( !$sensitive ) {
4829 # Remove trailing colon
4830 if ( substr( $syn, -1, 1 ) ===
':' ) {
4831 $syn = substr( $syn, 0, -1 );
4833 $this->mFunctionSynonyms[$sensitive][$syn] = $id;
4843 public function getFunctionHooks() {
4844 return array_keys( $this->mFunctionHooks );
4857 public function setFunctionTagHook( $tag, callable $callback,
$flags ) {
4858 $tag = strtolower( $tag );
4859 if ( preg_match(
'/[<>\r\n]/', $tag, $m ) ) {
4860 throw new MWException(
"Invalid character {$m[0]} in setFunctionTagHook('$tag', ...) call" );
4862 $old = isset( $this->mFunctionTagHooks[$tag] ) ?
4863 $this->mFunctionTagHooks[$tag] :
null;
4864 $this->mFunctionTagHooks[$tag] = [ $callback,
$flags ];
4866 if ( !in_array( $tag, $this->mStripList ) ) {
4867 $this->mStripList[] = $tag;
4880 public function replaceLinkHolders( &$text,
$options = 0 ) {
4881 $this->mLinkHolders->replace( $text );
4891 public function replaceLinkHoldersText( $text ) {
4892 return $this->mLinkHolders->replaceText( $text );
4908 public function renderImageGallery( $text,
$params ) {
4910 if ( isset(
$params[
'mode'] ) ) {
4916 }
catch ( Exception
$e ) {
4921 $ig->setContextTitle( $this->mTitle );
4922 $ig->setShowBytes(
false );
4923 $ig->setShowDimensions(
false );
4924 $ig->setShowFilename(
false );
4925 $ig->setParser( $this );
4926 $ig->setHideBadImages();
4927 $ig->setAttributes( Sanitizer::validateTagAttributes(
$params,
'ul' ) );
4929 if ( isset(
$params[
'showfilename'] ) ) {
4930 $ig->setShowFilename(
true );
4932 $ig->setShowFilename(
false );
4934 if ( isset(
$params[
'caption'] ) ) {
4935 $caption =
$params[
'caption'];
4936 $caption = htmlspecialchars( $caption );
4937 $caption = $this->replaceInternalLinks( $caption );
4938 $ig->setCaptionHtml( $caption );
4940 if ( isset(
$params[
'perrow'] ) ) {
4941 $ig->setPerRow(
$params[
'perrow'] );
4943 if ( isset(
$params[
'widths'] ) ) {
4944 $ig->setWidths(
$params[
'widths'] );
4946 if ( isset(
$params[
'heights'] ) ) {
4947 $ig->setHeights(
$params[
'heights'] );
4949 $ig->setAdditionalOptions(
$params );
4957 # match lines like these:
4958 # Image:someimage.jpg|This is some image
4966 if ( strpos(
$matches[0],
'%' ) !==
false ) {
4970 if ( is_null(
$title ) ) {
4971 # Bogus title. Ignore these so we don't bomb out later.
4975 # We need to get what handler the file uses, to figure out parameters.
4976 # Note, a hook can overide the file name, and chose an entirely different
4977 # file (which potentially could be of a different type and have different handler).
4982 # Don't register it now, as TraditionalImageGallery does that later.
4984 $handler = $file ? $file->getHandler() :
false;
4987 'img_alt' =>
'gallery-internal-alt',
4988 'img_link' =>
'gallery-internal-link',
4991 $paramMap = $paramMap +
$handler->getParamMap();
4994 unset( $paramMap[
'img_width'] );
5002 $handlerOptions = [];
5016 foreach ( $parameterMatches
as $parameterMatch ) {
5017 list( $magicName, $match ) = $mwArray->matchVariableStartToEnd( $parameterMatch );
5019 $paramName = $paramMap[$magicName];
5021 switch ( $paramName ) {
5022 case 'gallery-internal-alt':
5023 $alt = $this->stripAltText( $match,
false );
5025 case 'gallery-internal-link':
5026 $linkValue = strip_tags( $this->replaceLinkHoldersText( $match ) );
5027 $chars = self::EXT_LINK_URL_CLASS;
5028 $addr = self::EXT_LINK_ADDR;
5029 $prots = $this->mUrlProtocols;
5031 if ( preg_match(
'/^-{R|(.*)}-$/', $linkValue ) ) {
5034 $linkValue = substr( $linkValue, 4, -2 );
5036 if ( preg_match(
"/^($prots)$addr$chars*$/u", $linkValue ) ) {
5038 $this->mOutput->addExternalLink(
$link );
5041 if ( $localLinkTitle !==
null ) {
5042 $this->mOutput->addLink( $localLinkTitle );
5043 $link = $localLinkTitle->getLinkURL();
5049 if (
$handler->validateParam( $paramName, $match ) ) {
5050 $handlerOptions[$paramName] = $match;
5053 wfDebug(
"$parameterMatch failed parameter validation\n" );
5054 $label =
'|' . $parameterMatch;
5060 $label =
'|' . $parameterMatch;
5064 $label = substr( $label, 1 );
5067 $ig->add(
$title, $label, $alt,
$link, $handlerOptions );
5069 $html = $ig->toHTML();
5078 public function getImageParams(
$handler ) {
5080 $handlerClass = get_class(
$handler );
5084 if ( !isset( $this->mImageParams[$handlerClass] ) ) {
5085 # Initialise static lists
5086 static $internalParamNames = [
5087 'horizAlign' => [
'left',
'right',
'center',
'none' ],
5088 'vertAlign' => [
'baseline',
'sub',
'super',
'top',
'text-top',
'middle',
5089 'bottom',
'text-bottom' ],
5090 'frame' => [
'thumbnail',
'manualthumb',
'framed',
'frameless',
5091 'upright',
'border',
'link',
'alt',
'class' ],
5093 static $internalParamMap;
5094 if ( !$internalParamMap ) {
5095 $internalParamMap = [];
5096 foreach ( $internalParamNames
as $type => $names ) {
5103 $magicName = str_replace(
'-',
'_',
"img_$name" );
5104 $internalParamMap[$magicName] = [
$type,
$name ];
5109 # Add handler params
5110 $paramMap = $internalParamMap;
5112 $handlerParamMap =
$handler->getParamMap();
5113 foreach ( $handlerParamMap
as $magic => $paramName ) {
5114 $paramMap[$magic] = [
'handler', $paramName ];
5117 $this->mImageParams[$handlerClass] = $paramMap;
5118 $this->mImageParamsMagicArray[$handlerClass] =
new MagicWordArray( array_keys( $paramMap ) );
5120 return [ $this->mImageParams[$handlerClass], $this->mImageParamsMagicArray[$handlerClass] ];
5131 public function makeImage(
$title,
$options, $holders =
false ) {
5132 # Check if the options text is of the form "options|alt text"
5134 # * thumbnail make a thumbnail with enlarge-icon and caption, alignment depends on lang
5135 # * left no resizing, just left align. label is used for alt= only
5136 # * right same, but right aligned
5137 # * none same, but not aligned
5138 # * ___px scale to ___ pixels width, no aligning. e.g. use in taxobox
5139 # * center center the image
5140 # * frame Keep original image size, no magnify-button.
5141 # * framed Same as "frame"
5142 # * frameless like 'thumb' but without a frame. Keeps user preferences for width
5143 # * upright reduce width for upright images, rounded to full __0 px
5144 # * border draw a 1px border around the image
5145 # * alt Text for HTML alt attribute (defaults to empty)
5146 # * class Set a class for img node
5147 # * link Set the target of the image link. Can be external, interwiki, or local
5148 # vertical-align values (no % or length right now):
5158 # Protect LanguageConverter markup when splitting into parts
5163 # Give extensions a chance to select the file revision for us
5168 # Fetch and register the file (file title may be different via hooks)
5172 $handler = $file ? $file->getHandler() :
false;
5174 list( $paramMap, $mwArray ) = $this->getImageParams(
$handler );
5177 $this->addTrackingCategory(
'broken-file-category' );
5180 # Process the input parameters
5182 $params = [
'frame' => [],
'handler' => [],
5183 'horizAlign' => [],
'vertAlign' => [] ];
5184 $seenformat =
false;
5185 foreach ( $parts
as $part ) {
5186 $part = trim( $part );
5187 list( $magicName,
$value ) = $mwArray->matchVariableStartToEnd( $part );
5189 if ( isset( $paramMap[$magicName] ) ) {
5190 list(
$type, $paramName ) = $paramMap[$magicName];
5192 # Special case; width and height come in one variable together
5193 if (
$type ===
'handler' && $paramName ===
'width' ) {
5194 $parsedWidthParam = $this->parseWidthParam(
$value );
5195 if ( isset( $parsedWidthParam[
'width'] ) ) {
5196 $width = $parsedWidthParam[
'width'];
5197 if (
$handler->validateParam(
'width', $width ) ) {
5202 if ( isset( $parsedWidthParam[
'height'] ) ) {
5203 $height = $parsedWidthParam[
'height'];
5204 if (
$handler->validateParam(
'height', $height ) ) {
5209 # else no validation -- T15436
5211 if (
$type ===
'handler' ) {
5212 # Validate handler parameter
5215 # Validate internal parameters
5216 switch ( $paramName ) {
5220 # @todo FIXME: Possibly check validity here for
5221 # manualthumb? downstream behavior seems odd with
5222 # missing manual thumbs.
5227 $chars = self::EXT_LINK_URL_CLASS;
5228 $addr = self::EXT_LINK_ADDR;
5229 $prots = $this->mUrlProtocols;
5231 $paramName =
'no-link';
5234 } elseif ( preg_match(
"/^((?i)$prots)/",
$value ) ) {
5235 if ( preg_match(
"/^((?i)$prots)$addr$chars*$/u",
$value, $m ) ) {
5236 $paramName =
'link-url';
5237 $this->mOutput->addExternalLink(
$value );
5238 if ( $this->mOptions->getExternalLinkTarget() ) {
5239 $params[
$type][
'link-target'] = $this->mOptions->getExternalLinkTarget();
5246 $paramName =
'link-title';
5248 $this->mOutput->addLink( $linkTitle );
5257 $validated = !$seenformat;
5261 # Most other things appear to be empty or numeric...
5262 $validated = (
$value ===
false || is_numeric( trim(
$value ) ) );
5271 if ( !$validated ) {
5276 # Process alignment parameters
5277 if (
$params[
'horizAlign'] ) {
5284 $params[
'frame'][
'caption'] = $caption;
5286 # Will the image be presented in a frame, with the caption below?
5287 $imageIsFramed = isset(
$params[
'frame'][
'frame'] )
5288 || isset(
$params[
'frame'][
'framed'] )
5289 || isset(
$params[
'frame'][
'thumbnail'] )
5290 || isset(
$params[
'frame'][
'manualthumb'] );
5292 # In the old days, [[Image:Foo|text...]] would set alt text. Later it
5293 # came to also set the caption, ordinary text after the image -- which
5294 # makes no sense, because that just repeats the text multiple times in
5295 # screen readers. It *also* came to set the title attribute.
5296 # Now that we have an alt attribute, we should not set the alt text to
5297 # equal the caption: that's worse than useless, it just repeats the
5298 # text. This is the framed/thumbnail case. If there's no caption, we
5299 # use the unnamed parameter for alt text as well, just for the time be-
5300 # ing, if the unnamed param is set and the alt param is not.
5301 # For the future, we need to figure out if we want to tweak this more,
5302 # e.g., introducing a title= parameter for the title; ignoring the un-
5303 # named parameter entirely for images without a caption; adding an ex-
5304 # plicit caption= parameter and preserving the old magic unnamed para-
5306 if ( $imageIsFramed ) { # Framed image
5307 if ( $caption ===
'' && !isset(
$params[
'frame'][
'alt'] ) ) {
5308 # No caption or alt text, add the filename as the alt text so
5309 # that screen readers at least get some description of the image
5312 # Do not set $params['frame']['title'] because tooltips don't make sense
5314 }
else { # Inline image
5315 if ( !isset(
$params[
'frame'][
'alt'] ) ) {
5316 # No alt text, use the "caption" for the alt text
5317 if ( $caption !==
'' ) {
5318 $params[
'frame'][
'alt'] = $this->stripAltText( $caption, $holders );
5320 # No caption, fall back to using the filename for the
5325 # Use the "caption" for the tooltip text
5326 $params[
'frame'][
'title'] = $this->stripAltText( $caption, $holders );
5331 # Linker does the rest
5334 $time, $descQuery, $this->mOptions->getThumbSize() );
5336 # Give the handler a chance to modify the parser object
5338 $handler->parserTransformHook( $this, $file );
5349 protected function stripAltText( $caption, $holders ) {
5350 # Strip bad stuff out of the title (tooltip). We can't just use
5351 # replaceLinkHoldersText() here, because if this function is called
5352 # from replaceInternalLinks2(), mLinkHolders won't be up-to-date.
5354 $tooltip = $holders->replaceText( $caption );
5356 $tooltip = $this->replaceLinkHoldersText( $caption );
5359 # make sure there are no placeholders in thumbnail attributes
5360 # that are later expanded to html- so expand them now and
5362 $tooltip = $this->mStripState->unstripBoth( $tooltip );
5363 $tooltip = Sanitizer::stripAllTags( $tooltip );
5373 public function disableCache() {
5374 wfDebug(
"Parser output marked as uncacheable.\n" );
5375 if ( !$this->mOutput ) {
5377 " can only be called when actually parsing something" );
5379 $this->mOutput->updateCacheExpiry( 0 );
5390 public function attributeStripCallback( &$text, $frame =
false ) {
5391 $text = $this->replaceVariables( $text, $frame );
5392 $text = $this->mStripState->unstripBoth( $text );
5401 public function getTags() {
5403 array_keys( $this->mTransparentTagHooks ),
5404 array_keys( $this->mTagHooks ),
5405 array_keys( $this->mFunctionTagHooks )
5419 public function replaceTransparentTags( $text ) {
5421 $elements = array_keys( $this->mTransparentTagHooks );
5422 $text = self::extractTagsAndParams( $elements, $text,
$matches );
5426 list( $element, $content,
$params, $tag ) = $data;
5427 $tagName = strtolower( $element );
5428 if ( isset( $this->mTransparentTagHooks[$tagName] ) ) {
5429 $output = call_user_func_array(
5430 $this->mTransparentTagHooks[$tagName],
5436 $replacements[$marker] =
$output;
5438 return strtr( $text, $replacements );
5470 private function extractSections( $text, $sectionId, $mode, $newText =
'' ) {
5473 $magicScopeVariable = $this->lock();
5476 $frame = $this->getPreprocessor()->newFrame();
5478 # Process section extraction flags
5480 $sectionParts = explode(
'-', $sectionId );
5481 $sectionIndex = array_pop( $sectionParts );
5482 foreach ( $sectionParts
as $part ) {
5483 if ( $part ===
'T' ) {
5484 $flags |= self::PTD_FOR_INCLUSION;
5488 # Check for empty input
5489 if ( strval( $text ) ===
'' ) {
5490 # Only sections 0 and T-0 exist in an empty document
5491 if ( $sectionIndex == 0 ) {
5492 if ( $mode ===
'get' ) {
5498 if ( $mode ===
'get' ) {
5506 # Preprocess the text
5507 $root = $this->preprocessToDom( $text,
$flags );
5509 # <h> nodes indicate section breaks
5510 # They can only occur at the top level, so we can find them by iterating the root's children
5511 $node = $root->getFirstChild();
5513 # Find the target section
5514 if ( $sectionIndex == 0 ) {
5515 # Section zero doesn't nest, level=big
5516 $targetLevel = 1000;
5519 if ( $node->getName() ===
'h' ) {
5520 $bits = $node->splitHeading();
5521 if ( $bits[
'i'] == $sectionIndex ) {
5522 $targetLevel = $bits[
'level'];
5526 if ( $mode ===
'replace' ) {
5529 $node = $node->getNextSibling();
5535 if ( $mode ===
'get' ) {
5542 # Find the end of the section, including nested sections
5544 if ( $node->getName() ===
'h' ) {
5545 $bits = $node->splitHeading();
5546 $curLevel = $bits[
'level'];
5547 if ( $bits[
'i'] != $sectionIndex && $curLevel <= $targetLevel ) {
5551 if ( $mode ===
'get' ) {
5554 $node = $node->getNextSibling();
5557 # Write out the remainder (in replace mode only)
5558 if ( $mode ===
'replace' ) {
5559 # Output the replacement text
5560 # Add two newlines on -- trailing whitespace in $newText is conventionally
5561 # stripped by the editor, so we need both newlines to restore the paragraph gap
5562 # Only add trailing whitespace if there is newText
5563 if ( $newText !=
"" ) {
5564 $outText .= $newText .
"\n\n";
5569 $node = $node->getNextSibling();
5573 if ( is_string( $outText ) ) {
5574 # Re-insert stripped tags
5575 $outText = rtrim( $this->mStripState->unstripBoth( $outText ) );
5595 public function getSection( $text, $sectionId, $defaultText =
'' ) {
5596 return $this->extractSections( $text, $sectionId,
'get', $defaultText );
5611 public function replaceSection( $oldText, $sectionId, $newText ) {
5612 return $this->extractSections( $oldText, $sectionId,
'replace', $newText );
5620 public function getRevisionId() {
5621 return $this->mRevisionId;
5630 public function getRevisionObject() {
5631 if ( !is_null( $this->mRevisionObject ) ) {
5632 return $this->mRevisionObject;
5634 if ( is_null( $this->mRevisionId ) ) {
5638 $rev = call_user_func(
5639 $this->mOptions->getCurrentRevisionCallback(), $this->getTitle(), $this
5642 # If the parse is for a new revision, then the callback should have
5643 # already been set to force the object and should match mRevisionId.
5644 # If not, try to fetch by mRevisionId for sanity.
5645 if (
$rev &&
$rev->getId() != $this->mRevisionId ) {
5649 $this->mRevisionObject =
$rev;
5651 return $this->mRevisionObject;
5659 public function getRevisionTimestamp() {
5660 if ( is_null( $this->mRevisionTimestamp ) ) {
5663 $revObject = $this->getRevisionObject();
5664 $timestamp = $revObject ? $revObject->getTimestamp() :
wfTimestampNow();
5666 # The cryptic '' timezone parameter tells to use the site-default
5667 # timezone offset instead of the user settings.
5668 # Since this value will be saved into the parser cache, served
5669 # to other users, and potentially even used inside links and such,
5670 # it needs to be consistent for all visitors.
5671 $this->mRevisionTimestamp =
$wgContLang->userAdjust( $timestamp,
'' );
5674 return $this->mRevisionTimestamp;
5682 public function getRevisionUser() {
5683 if ( is_null( $this->mRevisionUser ) ) {
5684 $revObject = $this->getRevisionObject();
5686 # if this template is subst: the revision id will be blank,
5687 # so just use the current user's name
5689 $this->mRevisionUser = $revObject->getUserText();
5690 } elseif ( $this->ot[
'wiki'] || $this->mOptions->getIsPreview() ) {
5691 $this->mRevisionUser = $this->getUser()->getName();
5694 return $this->mRevisionUser;
5702 public function getRevisionSize() {
5703 if ( is_null( $this->mRevisionSize ) ) {
5704 $revObject = $this->getRevisionObject();
5706 # if this variable is subst: the revision id will be blank,
5707 # so just use the parser input size, because the own substituation
5708 # will change the size.
5710 $this->mRevisionSize = $revObject->getSize();
5712 $this->mRevisionSize = $this->mInputSize;
5715 return $this->mRevisionSize;
5723 public function setDefaultSort(
$sort ) {
5724 $this->mDefaultSort =
$sort;
5725 $this->mOutput->setProperty(
'defaultsort',
$sort );
5738 public function getDefaultSort() {
5739 if ( $this->mDefaultSort !==
false ) {
5740 return $this->mDefaultSort;
5752 public function getCustomDefaultSort() {
5753 return $this->mDefaultSort;
5765 public function guessSectionNameFromWikiText( $text ) {
5766 # Strip out wikitext links(they break the anchor)
5767 $text = $this->stripSectionName( $text );
5768 $text = Sanitizer::normalizeSectionNameWhitespace( $text );
5769 $text = Sanitizer::decodeCharReferences( $text );
5770 return '#' . Sanitizer::escapeIdForLink( $text );
5782 public function guessLegacySectionNameFromWikiText( $text ) {
5785 # Strip out wikitext links(they break the anchor)
5786 $text = $this->stripSectionName( $text );
5787 $text = Sanitizer::normalizeSectionNameWhitespace( $text );
5788 $text = Sanitizer::decodeCharReferences( $text );
5792 $id = Sanitizer::escapeIdForAttribute( $text, Sanitizer::ID_FALLBACK );
5794 $id = Sanitizer::escapeIdForLink( $text );
5814 public function stripSectionName( $text ) {
5815 # Strip internal link markup
5816 $text = preg_replace(
'/\[\[:?([^[|]+)\|([^[]+)\]\]/',
'$2', $text );
5817 $text = preg_replace(
'/\[\[:?([^[]+)\|?\]\]/',
'$1', $text );
5819 # Strip external link markup
5820 # @todo FIXME: Not tolerant to blank link text
5821 # I.E. [https://www.mediawiki.org] will render as [1] or something depending
5822 # on how many empty links there are on the page - need to figure that out.
5823 $text = preg_replace(
'/\[(?i:' . $this->mUrlProtocols .
')([^ ]+?) ([^[]+)\]/',
'$2', $text );
5825 # Parse wikitext quotes (italics & bold)
5826 $text = $this->doQuotes( $text );
5846 $magicScopeVariable = $this->lock();
5849 $text = $this->replaceVariables( $text );
5850 $text = $this->mStripState->unstripBoth( $text );
5851 $text = Sanitizer::removeHTMLtags( $text );
5891 public function markerSkipCallback(
$s, $callback ) {
5894 while ( $i < strlen(
$s ) ) {
5895 $markerStart = strpos(
$s, self::MARKER_PREFIX, $i );
5896 if ( $markerStart ===
false ) {
5897 $out .= call_user_func( $callback, substr(
$s, $i ) );
5900 $out .= call_user_func( $callback, substr(
$s, $i, $markerStart - $i ) );
5901 $markerEnd = strpos(
$s, self::MARKER_SUFFIX, $markerStart );
5902 if ( $markerEnd ===
false ) {
5903 $out .= substr(
$s, $markerStart );
5906 $markerEnd += strlen( self::MARKER_SUFFIX );
5907 $out .= substr(
$s, $markerStart, $markerEnd - $markerStart );
5921 public function killMarkers( $text ) {
5922 return $this->mStripState->killMarkers( $text );
5941 public function serializeHalfParsedText( $text ) {
5944 'version' => self::HALF_PARSED_VERSION,
5945 'stripState' => $this->mStripState->getSubState( $text ),
5946 'linkHolders' => $this->mLinkHolders->getSubArray( $text )
5966 public function unserializeHalfParsedText( $data ) {
5967 if ( !isset( $data[
'version'] ) || $data[
'version'] != self::HALF_PARSED_VERSION ) {
5968 throw new MWException( __METHOD__ .
': invalid version' );
5971 # First, extract the strip state.
5972 $texts = [ $data[
'text'] ];
5973 $texts = $this->mStripState->merge( $data[
'stripState'], $texts );
5975 # Now renumber links
5976 $texts = $this->mLinkHolders->mergeForeign( $data[
'linkHolders'], $texts );
5978 # Should be good to go.
5991 public function isValidHalfParsedText( $data ) {
5992 return isset( $data[
'version'] ) && $data[
'version'] == self::HALF_PARSED_VERSION;
6003 public function parseWidthParam(
$value ) {
6004 $parsedWidthParam = [];
6006 return $parsedWidthParam;
6009 # (T15500) In both cases (width/height and width only),
6010 # permit trailing "px" for backward compatibility.
6011 if ( preg_match(
'/^([0-9]*)x([0-9]*)\s*(?:px)?\s*$/',
$value, $m ) ) {
6012 $width = intval( $m[1] );
6013 $height = intval( $m[2] );
6014 $parsedWidthParam[
'width'] = $width;
6015 $parsedWidthParam[
'height'] = $height;
6016 } elseif ( preg_match(
'/^[0-9]*\s*(?:px)?\s*$/',
$value ) ) {
6017 $width = intval(
$value );
6018 $parsedWidthParam[
'width'] = $width;
6020 return $parsedWidthParam;
6032 protected function lock() {
6033 if ( $this->mInParse ) {
6034 throw new MWException(
"Parser state cleared while parsing. "
6035 .
"Did you call Parser::parse recursively? Lock is held by: " . $this->mInParse );
6041 $this->mInParse =
$e->getTraceAsString();
6043 $recursiveCheck =
new ScopedCallback(
function () {
6044 $this->mInParse =
false;
6047 return $recursiveCheck;
6060 public static function stripOuterParagraph(
$html ) {
6062 if ( preg_match(
'/^<p>(.*)\n?<\/p>\n?$/sU',
$html, $m ) ) {
6063 if ( strpos( $m[1],
'</p>' ) ===
false ) {
6081 public function getFreshParser() {
6083 if ( $this->mInParse ) {
6096 public function enableOOUI() {
6098 $this->mOutput->setEnableOOUI(
true );