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 = Parser::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";
525 if ( $wgShowHostnames ) {
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() ) {
542 $valueMsg =
new RawMessage(
'$1' );
544 if ( !$keyMsg->isDisabled() && !$valueMsg->isDisabled() ) {
545 $valueMsg->params(
$value );
546 $limitReport .=
"{$keyMsg->text()}: {$valueMsg->text()}\n";
552 $limitReport = htmlspecialchars_decode( $limitReport );
553 Hooks::run(
'ParserLimitReport', [ $this, &$limitReport ] );
557 $limitReport = str_replace( [
'-',
'&' ], [
'‐',
'&' ], $limitReport );
558 $text .=
"\n<!-- \n$limitReport-->\n";
561 $dataByFunc = $this->mProfiler->getFunctionStats();
562 uasort( $dataByFunc,
function ( $a, $b ) {
563 return $a[
'real'] < $b[
'real'];
566 foreach ( array_slice( $dataByFunc, 0, 10 )
as $item ) {
567 $profileReport[] = sprintf(
"%6.2f%% %8.3f %6d %s",
568 $item[
'%real'], $item[
'real'], $item[
'calls'],
569 htmlspecialchars( $item[
'name'] ) );
571 $text .=
"<!--\nTransclusion expansion time report (%,ms,calls,template)\n";
572 $text .= implode(
"\n", $profileReport ) .
"\n-->\n";
574 $this->mOutput->setLimitReportData(
'limitreport-timingprofile', $profileReport );
577 if ( $wgShowHostnames ) {
578 $this->mOutput->setLimitReportData(
'cachereport-origin',
wfHostname() );
580 $this->mOutput->setLimitReportData(
'cachereport-timestamp',
581 $this->mOutput->getCacheTime() );
582 $this->mOutput->setLimitReportData(
'cachereport-ttl',
583 $this->mOutput->getCacheExpiry() );
584 $this->mOutput->setLimitReportData(
'cachereport-transientcontent',
585 $this->mOutput->hasDynamicContent() );
587 if ( $this->mGeneratedPPNodeCount > $this->mOptions->getMaxGeneratedPPNodeCount() / 10 ) {
588 wfDebugLog(
'generated-pp-node-count', $this->mGeneratedPPNodeCount .
' ' .
589 $this->mTitle->getPrefixedDBkey() );
592 $this->mOutput->setText( $text );
594 $this->mRevisionId = $oldRevisionId;
595 $this->mRevisionObject = $oldRevisionObject;
596 $this->mRevisionTimestamp = $oldRevisionTimestamp;
597 $this->mRevisionUser = $oldRevisionUser;
598 $this->mRevisionSize = $oldRevisionSize;
599 $this->mInputSize =
false;
600 $this->currentRevisionCache =
null;
602 return $this->mOutput;
627 public function recursiveTagParse( $text, $frame =
false ) {
632 $text = $this->internalParse( $text,
false, $frame );
653 public function recursiveTagParseFully( $text, $frame =
false ) {
654 $text = $this->recursiveTagParse( $text, $frame );
655 $text = $this->internalParseHalfParsed( $text,
false );
670 public function preprocess( $text,
Title $title =
null,
673 $magicScopeVariable = $this->lock();
675 if ( $revid !==
null ) {
676 $this->mRevisionId = $revid;
682 $text = $this->replaceVariables( $text, $frame );
683 $text = $this->mStripState->unstripBoth( $text );
696 public function recursivePreprocess( $text, $frame =
false ) {
697 $text = $this->replaceVariables( $text, $frame );
698 $text = $this->mStripState->unstripBoth( $text );
716 $msg =
new RawMessage( $text );
717 $text = $msg->params(
$params )->plain();
719 # Parser (re)initialisation
720 $magicScopeVariable = $this->lock();
724 $dom = $this->preprocessToDom( $text, self::PTD_FOR_INCLUSION );
725 $text = $this->getPreprocessor()->newFrame()->expand( $dom,
$flags );
726 $text = $this->mStripState->unstripBoth( $text );
736 public static function getRandomString() {
747 public function setUser(
$user ) {
748 $this->mUser =
$user;
757 public function uniqPrefix() {
759 return self::MARKER_PREFIX;
767 public function setTitle(
$t ) {
772 if (
$t->hasFragment() ) {
773 # Strip the fragment to avoid various odd effects
774 $this->mTitle =
$t->createFragmentTarget(
'' );
785 public function getTitle() {
786 return $this->mTitle;
795 public function Title( $x =
null ) {
796 return wfSetVar( $this->mTitle, $x );
804 public function setOutputType( $ot ) {
805 $this->mOutputType = $ot;
821 public function OutputType( $x =
null ) {
822 return wfSetVar( $this->mOutputType, $x );
830 public function getOutput() {
831 return $this->mOutput;
839 public function getOptions() {
840 return $this->mOptions;
849 public function Options( $x =
null ) {
850 return wfSetVar( $this->mOptions, $x );
856 public function nextLinkID() {
857 return $this->mLinkID++;
863 public function setLinkID( $id ) {
864 $this->mLinkID = $id;
871 public function getFunctionLang() {
872 return $this->getTargetLanguage();
884 public function getTargetLanguage() {
885 $target = $this->mOptions->getTargetLanguage();
887 if ( $target !==
null ) {
889 } elseif ( $this->mOptions->getInterfaceMessage() ) {
890 return $this->mOptions->getUserLangObj();
891 } elseif ( is_null( $this->mTitle ) ) {
892 throw new MWException( __METHOD__ .
': $this->mTitle is null' );
895 return $this->mTitle->getPageLanguage();
902 public function getConverterLanguage() {
903 return $this->getTargetLanguage();
912 public function getUser() {
913 if ( !is_null( $this->mUser ) ) {
916 return $this->mOptions->getUser();
924 public function getPreprocessor() {
925 if ( !isset( $this->mPreprocessor ) ) {
926 $class = $this->mPreprocessorClass;
927 $this->mPreprocessor =
new $class( $this );
929 return $this->mPreprocessor;
938 public function getLinkRenderer() {
939 if ( !$this->mLinkRenderer ) {
940 $this->mLinkRenderer = MediaWikiServices::getInstance()
941 ->getLinkRendererFactory()->create();
942 $this->mLinkRenderer->setStubThreshold(
943 $this->getOptions()->getStubThreshold()
947 return $this->mLinkRenderer;
971 public static function extractTagsAndParams( $elements, $text, &
$matches, $uniq_prefix =
null ) {
972 if ( $uniq_prefix !==
null ) {
973 wfDeprecated( __METHOD__ .
' called with $prefix argument',
'1.26' );
979 $taglist = implode(
'|', $elements );
980 $start =
"/<($taglist)(\\s+[^>]*?|\\s*?)(\/?" .
">)|<(!--)/i";
982 while ( $text !=
'' ) {
983 $p = preg_split( $start, $text, 2, PREG_SPLIT_DELIM_CAPTURE );
985 if (
count( $p ) < 5 ) {
988 if (
count( $p ) > 5 ) {
1002 $marker = self::MARKER_PREFIX .
"-$element-" . sprintf(
'%08X', $n++ ) . self::MARKER_SUFFIX;
1003 $stripped .= $marker;
1005 if ( $close ===
'/>' ) {
1006 # Empty element tag, <tag />
1011 if ( $element ===
'!--' ) {
1014 $end =
"/(<\\/$element\\s*>)/i";
1016 $q = preg_split( $end, $inside, 2, PREG_SPLIT_DELIM_CAPTURE );
1018 if (
count( $q ) < 3 ) {
1019 # No end tag -- let it run out to the end of the text.
1030 Sanitizer::decodeTagAttributes( $attributes ),
1031 "<$element$attributes$close$content$tail" ];
1041 public function getStripList() {
1042 return $this->mStripList;
1054 public function insertStripItem( $text ) {
1055 $marker = self::MARKER_PREFIX .
"-item-{$this->mMarkerIndex}-" . self::MARKER_SUFFIX;
1056 $this->mMarkerIndex++;
1057 $this->mStripState->addGeneral( $marker, $text );
1068 public function doTableStuff( $text ) {
1072 $td_history = []; # Is currently
a td
tag open?
1073 $last_tag_history = []; # Save
history of last lag activated (td, th
or caption)
1074 $tr_history = []; # Is currently
a tr
tag open?
1075 $tr_attributes = []; #
history of tr attributes
1076 $has_opened_tr = []; # Did
this table open a <tr> element?
1077 $indent_level = 0; # indent level
of the
table
1080 $line = trim( $outLine );
1083 $out .= $outLine .
"\n";
1087 $first_character =
$line[0];
1088 $first_two = substr(
$line, 0, 2 );
1091 if ( preg_match(
'/^(:*)\s*\{\|(.*)$/',
$line,
$matches ) ) {
1092 # First check if we are starting a new table
1093 $indent_level = strlen(
$matches[1] );
1095 $attributes = $this->mStripState->unstripBoth(
$matches[2] );
1096 $attributes = Sanitizer::fixTagAttributes( $attributes,
'table' );
1098 $outLine = str_repeat(
'<dl><dd>', $indent_level ) .
"<table{$attributes}>";
1099 array_push( $td_history,
false );
1100 array_push( $last_tag_history,
'' );
1101 array_push( $tr_history,
false );
1102 array_push( $tr_attributes,
'' );
1103 array_push( $has_opened_tr,
false );
1104 } elseif (
count( $td_history ) == 0 ) {
1105 # Don't do any of the following
1106 $out .= $outLine .
"\n";
1108 } elseif ( $first_two ===
'|}' ) {
1109 # We are ending a table
1111 $last_tag = array_pop( $last_tag_history );
1113 if ( !array_pop( $has_opened_tr ) ) {
1114 $line =
"<tr><td></td></tr>{$line}";
1117 if ( array_pop( $tr_history ) ) {
1118 $line =
"</tr>{$line}";
1121 if ( array_pop( $td_history ) ) {
1122 $line =
"</{$last_tag}>{$line}";
1124 array_pop( $tr_attributes );
1125 $outLine =
$line . str_repeat(
'</dd></dl>', $indent_level );
1126 } elseif ( $first_two ===
'|-' ) {
1127 # Now we have a table row
1128 $line = preg_replace(
'#^\|-+#',
'',
$line );
1130 # Whats after the tag is now only attributes
1131 $attributes = $this->mStripState->unstripBoth(
$line );
1132 $attributes = Sanitizer::fixTagAttributes( $attributes,
'tr' );
1133 array_pop( $tr_attributes );
1134 array_push( $tr_attributes, $attributes );
1137 $last_tag = array_pop( $last_tag_history );
1138 array_pop( $has_opened_tr );
1139 array_push( $has_opened_tr,
true );
1141 if ( array_pop( $tr_history ) ) {
1145 if ( array_pop( $td_history ) ) {
1146 $line =
"</{$last_tag}>{$line}";
1150 array_push( $tr_history,
false );
1151 array_push( $td_history,
false );
1152 array_push( $last_tag_history,
'' );
1153 } elseif ( $first_character ===
'|'
1154 || $first_character ===
'!'
1155 || $first_two ===
'|+'
1157 # This might be cell elements, td, th or captions
1158 if ( $first_two ===
'|+' ) {
1159 $first_character =
'+';
1166 if ( $first_character ===
'!' ) {
1170 # Split up multiple cells on the same line.
1171 # FIXME : This can result in improper nesting of tags processed
1172 # by earlier parser steps.
1173 $cells = explode(
'||',
$line );
1177 # Loop through each table cell
1178 foreach ( $cells
as $cell ) {
1180 if ( $first_character !==
'+' ) {
1181 $tr_after = array_pop( $tr_attributes );
1182 if ( !array_pop( $tr_history ) ) {
1183 $previous =
"<tr{$tr_after}>\n";
1185 array_push( $tr_history,
true );
1186 array_push( $tr_attributes,
'' );
1187 array_pop( $has_opened_tr );
1188 array_push( $has_opened_tr,
true );
1191 $last_tag = array_pop( $last_tag_history );
1193 if ( array_pop( $td_history ) ) {
1194 $previous =
"</{$last_tag}>\n{$previous}";
1197 if ( $first_character ===
'|' ) {
1199 } elseif ( $first_character ===
'!' ) {
1201 } elseif ( $first_character ===
'+' ) {
1202 $last_tag =
'caption';
1207 array_push( $last_tag_history, $last_tag );
1209 # A cell could contain both parameters and data
1210 $cell_data = explode(
'|', $cell, 2 );
1212 # T2553: Note that a '|' inside an invalid link should not
1213 # be mistaken as delimiting cell parameters
1214 # Bug T153140: Neither should language converter markup.
1215 if ( preg_match(
'/\[\[|-\{/', $cell_data[0] ) === 1 ) {
1216 $cell =
"{$previous}<{$last_tag}>{$cell}";
1217 } elseif (
count( $cell_data ) == 1 ) {
1218 $cell =
"{$previous}<{$last_tag}>{$cell_data[0]}";
1220 $attributes = $this->mStripState->unstripBoth( $cell_data[0] );
1221 $attributes = Sanitizer::fixTagAttributes( $attributes, $last_tag );
1222 $cell =
"{$previous}<{$last_tag}{$attributes}>{$cell_data[1]}";
1226 array_push( $td_history,
true );
1229 $out .= $outLine .
"\n";
1232 # Closing open td, tr && table
1233 while (
count( $td_history ) > 0 ) {
1234 if ( array_pop( $td_history ) ) {
1237 if ( array_pop( $tr_history ) ) {
1240 if ( !array_pop( $has_opened_tr ) ) {
1241 $out .=
"<tr><td></td></tr>\n";
1244 $out .=
"</table>\n";
1247 # Remove trailing line-ending (b/c)
1248 if ( substr(
$out, -1 ) ===
"\n" ) {
1252 # special case: don't return empty table
1253 if (
$out ===
"<table>\n<tr><td></td></tr>\n</table>" ) {
1272 public function internalParse( $text, $isMain =
true, $frame =
false ) {
1279 # Hook to suspend the parser in this state
1280 if ( !
Hooks::run(
'ParserBeforeInternalParse', [ &
$parser, &$text, &$this->mStripState ] ) ) {
1284 # if $frame is provided, then use $frame for replacing any variables
1286 # use frame depth to infer how include/noinclude tags should be handled
1287 # depth=0 means this is the top-level document; otherwise it's an included document
1288 if ( !$frame->depth ) {
1291 $flag = Parser::PTD_FOR_INCLUSION;
1293 $dom = $this->preprocessToDom( $text, $flag );
1294 $text = $frame->expand( $dom );
1296 # if $frame is not provided, then use old-style replaceVariables
1297 $text = $this->replaceVariables( $text );
1300 Hooks::run(
'InternalParseBeforeSanitize', [ &
$parser, &$text, &$this->mStripState ] );
1301 $text = Sanitizer::removeHTMLtags(
1303 [ $this,
'attributeStripCallback' ],
1305 array_keys( $this->mTransparentTagHooks ),
1307 [ $this,
'addTrackingCategory' ]
1309 Hooks::run(
'InternalParseBeforeLinks', [ &
$parser, &$text, &$this->mStripState ] );
1311 # Tables need to come after variable replacement for things to work
1312 # properly; putting them before other transformations should keep
1313 # exciting things like link expansions from showing up in surprising
1315 $text = $this->doTableStuff( $text );
1317 $text = preg_replace(
'/(^|\n)-----*/',
'\\1<hr />', $text );
1319 $text = $this->doDoubleUnderscore( $text );
1321 $text = $this->doHeadings( $text );
1322 $text = $this->replaceInternalLinks( $text );
1323 $text = $this->doAllQuotes( $text );
1324 $text = $this->replaceExternalLinks( $text );
1326 # replaceInternalLinks may sometimes leave behind
1327 # absolute URLs, which have to be masked to hide them from replaceExternalLinks
1328 $text = str_replace( self::MARKER_PREFIX .
'NOPARSE',
'', $text );
1330 $text = $this->doMagicLinks( $text );
1331 $text = $this->formatHeadings( $text, $origText, $isMain );
1345 private function internalParseHalfParsed( $text, $isMain =
true, $linestart =
true ) {
1346 $text = $this->mStripState->unstripGeneral( $text );
1355 # Clean up special characters, only run once, next-to-last before doBlockLevels
1357 # French spaces, last one Guillemet-left
1358 # only if there is something before the space
1359 '/(.) (?=\\?|:|;|!|%|\\302\\273)/' =>
'\\1 ',
1360 # french spaces, Guillemet-right
1361 '/(\\302\\253) /' =>
'\\1 ',
1362 '/ (!\s*important)/' =>
' \\1', # Beware
of CSS magic
word !important, T13874.
1364 $text = preg_replace( array_keys( $fixtags ), array_values( $fixtags ), $text );
1366 $text = $this->doBlockLevels( $text, $linestart );
1368 $this->replaceLinkHolders( $text );
1377 if ( !( $this->mOptions->getDisableContentConversion()
1378 || isset( $this->mDoubleUnderscores[
'nocontentconvert'] ) )
1380 if ( !$this->mOptions->getInterfaceMessage() ) {
1381 # The position of the convert() call should not be changed. it
1382 # assumes that the links are all replaced and the only thing left
1383 # is the <nowiki> mark.
1384 $text = $this->getConverterLanguage()->convert( $text );
1388 $text = $this->mStripState->unstripNoWiki( $text );
1394 $text = $this->replaceTransparentTags( $text );
1395 $text = $this->mStripState->unstripGeneral( $text );
1397 $text = Sanitizer::normalizeCharReferences( $text );
1400 if ( $this->mOptions->getTidy() ) {
1404 # attempt to sanitize at least some nesting problems
1405 # (T4702 and quite a few others)
1407 # ''Something [http://www.cool.com cool''] -->
1408 # <i>Something</i><a href="http://www.cool.com"..><i>cool></i></a>
1409 '/(<([bi])>)(<([bi])>)?([^<]*)(<\/?a[^<]*>)([^<]*)(<\/\\4>)?(<\/\\2>)/' =>
1410 '\\1\\3\\5\\8\\9\\6\\1\\3\\7\\8\\9',
1411 # fix up an anchor inside another anchor, only
1412 # at least for a single single nested link (T5695)
1413 '/(<a[^>]+>)([^<]*)(<a[^>]+>[^<]*)<\/a>(.*)<\/a>/' =>
1414 '\\1\\2</a>\\3</a>\\1\\4</a>',
1415 # fix div inside inline elements- doBlockLevels won't wrap a line which
1416 # contains a div, so fix it up here; replace
1417 # div with escaped text
1418 '/(<([aib]) [^>]+>)([^<]*)(<div([^>]*)>)(.*)(<\/div>)([^<]*)(<\/\\2>)/' =>
1419 '\\1\\3<div\\5>\\6</div>\\8\\9',
1420 # remove empty italic or bold tag pairs, some
1421 # introduced by rules above
1422 '/<([bi])><\/\\1>/' =>
'',
1425 $text = preg_replace(
1426 array_keys( $tidyregs ),
1427 array_values( $tidyregs ),
1449 public function doMagicLinks( $text ) {
1451 $urlChar = self::EXT_LINK_URL_CLASS;
1452 $addr = self::EXT_LINK_ADDR;
1453 $space = self::SPACE_NOT_NL; # non-newline space
1454 $spdash =
"(?:-|$space)"; #
a dash
or a non-newline space
1455 $spaces =
"$space++"; # possessive match
of 1
or more spaces
1456 $text = preg_replace_callback(
1458 (<a[ \t\r\n>].*?</a>) | # m[1]: Skip link text
1459 (<.*?>) | # m[2]: Skip stuff inside HTML elements' .
"
1460 (\b # m[3]: Free external links
1462 ($addr$urlChar*) # m[4]: Post-protocol path
1464 \b(?:RFC|PMID) $spaces # m[5]: RFC or PMID, capture number
1466 \bISBN $spaces ( # m[6]: ISBN, capture number
1467 (?: 97[89] $spdash? )? # optional 13-digit ISBN prefix
1468 (?: [0-9] $spdash? ){9} # 9 digits with opt. delimiters
1469 [0-9Xx] # check digit
1471 )!xu", [ $this,
'magicLinkCallback' ], $text );
1480 public function magicLinkCallback( $m ) {
1481 if ( isset( $m[1] ) && $m[1] !==
'' ) {
1484 } elseif ( isset( $m[2] ) && $m[2] !==
'' ) {
1487 } elseif ( isset( $m[3] ) && $m[3] !==
'' ) {
1488 # Free external link
1489 return $this->makeFreeExternalLink( $m[0], strlen( $m[4] ) );
1490 } elseif ( isset( $m[5] ) && $m[5] !==
'' ) {
1492 if ( substr( $m[0], 0, 3 ) ===
'RFC' ) {
1493 if ( !$this->mOptions->getMagicRFCLinks() ) {
1498 $cssClass =
'mw-magiclink-rfc';
1499 $trackingCat =
'magiclink-tracking-rfc';
1501 } elseif ( substr( $m[0], 0, 4 ) ===
'PMID' ) {
1502 if ( !$this->mOptions->getMagicPMIDLinks() ) {
1506 $urlmsg =
'pubmedurl';
1507 $cssClass =
'mw-magiclink-pmid';
1508 $trackingCat =
'magiclink-tracking-pmid';
1511 throw new MWException( __METHOD__ .
': unrecognised match type "' .
1512 substr( $m[0], 0, 20 ) .
'"' );
1514 $url =
wfMessage( $urlmsg, $id )->inContentLanguage()->text();
1515 $this->addTrackingCategory( $trackingCat );
1517 } elseif ( isset( $m[6] ) && $m[6] !==
''
1518 && $this->mOptions->getMagicISBNLinks()
1522 $space = self::SPACE_NOT_NL; # non-newline space
1523 $isbn = preg_replace(
"/$space/",
' ', $isbn );
1524 $num = strtr( $isbn, [
1529 $this->addTrackingCategory(
'magiclink-tracking-isbn' );
1530 return $this->getLinkRenderer()->makeKnownLink(
1534 'class' =>
'internal mw-magiclink-isbn',
1552 public function makeFreeExternalLink( $url, $numPostProto ) {
1555 # The characters '<' and '>' (which were escaped by
1556 # removeHTMLtags()) should not be included in
1557 # URLs, per RFC 2396.
1558 # Make terminate a URL as well (bug T84937)
1561 '/&(lt|gt|nbsp|#x0*(3[CcEe]|[Aa]0)|#0*(60|62|160));/',
1566 $trail = substr( $url, $m2[0][1] ) . $trail;
1567 $url = substr( $url, 0, $m2[0][1] );
1570 # Move trailing punctuation to $trail
1572 # If there is no left bracket, then consider right brackets fair game too
1573 if ( strpos( $url,
'(' ) ===
false ) {
1577 $urlRev = strrev( $url );
1578 $numSepChars = strspn( $urlRev, $sep );
1579 # Don't break a trailing HTML entity by moving the ; into $trail
1580 # This is in hot code, so use substr_compare to avoid having to
1581 # create a new string object for the comparison
1582 if ( $numSepChars && substr_compare( $url,
";", -$numSepChars, 1 ) === 0 ) {
1583 # more optimization: instead of running preg_match with a $
1584 # anchor, which can be slow, do the match on the reversed
1585 # string starting at the desired offset.
1586 # un-reversed regexp is: /&([a-z]+|#x[\da-f]+|#\d+)$/i
1587 if ( preg_match(
'/\G([a-z]+|[\da-f]+x#|\d+#)&/i', $urlRev, $m2, 0, $numSepChars ) ) {
1591 if ( $numSepChars ) {
1592 $trail = substr( $url, -$numSepChars ) . $trail;
1593 $url = substr( $url, 0, -$numSepChars );
1596 # Verify that we still have a real URL after trail removal, and
1597 # not just lone protocol
1598 if ( strlen( $trail ) >= $numPostProto ) {
1599 return $url . $trail;
1602 $url = Sanitizer::cleanUrl( $url );
1604 # Is this an external image?
1605 $text = $this->maybeMakeExternalImage( $url );
1606 if ( $text ===
false ) {
1607 # Not an image, make a link
1609 $this->getConverterLanguage()->markNoConversion( $url,
true ),
1611 $this->getExternalLinkAttribs( $url ), $this->mTitle );
1612 # Register it in the output object...
1613 $this->mOutput->addExternalLink( $url );
1615 return $text . $trail;
1627 public function doHeadings( $text ) {
1628 for ( $i = 6; $i >= 1; --$i ) {
1629 $h = str_repeat(
'=', $i );
1630 $text = preg_replace(
"/^$h(.+)$h\\s*$/m",
"<h$i>\\1</h$i>", $text );
1643 public function doAllQuotes( $text ) {
1647 $outtext .= $this->doQuotes(
$line ) .
"\n";
1649 $outtext = substr( $outtext, 0, -1 );
1660 public function doQuotes( $text ) {
1661 $arr = preg_split(
"/(''+)/", $text, -1, PREG_SPLIT_DELIM_CAPTURE );
1662 $countarr =
count( $arr );
1663 if ( $countarr == 1 ) {
1672 for ( $i = 1; $i < $countarr; $i += 2 ) {
1673 $thislen = strlen( $arr[$i] );
1677 if ( $thislen == 4 ) {
1678 $arr[$i - 1] .=
"'";
1681 } elseif ( $thislen > 5 ) {
1685 $arr[$i - 1] .= str_repeat(
"'", $thislen - 5 );
1690 if ( $thislen == 2 ) {
1692 } elseif ( $thislen == 3 ) {
1694 } elseif ( $thislen == 5 ) {
1704 if ( ( $numbold % 2 == 1 ) && ( $numitalics % 2 == 1 ) ) {
1705 $firstsingleletterword = -1;
1706 $firstmultiletterword = -1;
1708 for ( $i = 1; $i < $countarr; $i += 2 ) {
1709 if ( strlen( $arr[$i] ) == 3 ) {
1710 $x1 = substr( $arr[$i - 1], -1 );
1711 $x2 = substr( $arr[$i - 1], -2, 1 );
1712 if ( $x1 ===
' ' ) {
1713 if ( $firstspace == -1 ) {
1716 } elseif ( $x2 ===
' ' ) {
1717 $firstsingleletterword = $i;
1722 if ( $firstmultiletterword == -1 ) {
1723 $firstmultiletterword = $i;
1730 if ( $firstsingleletterword > -1 ) {
1731 $arr[$firstsingleletterword] =
"''";
1732 $arr[$firstsingleletterword - 1] .=
"'";
1733 } elseif ( $firstmultiletterword > -1 ) {
1735 $arr[$firstmultiletterword] =
"''";
1736 $arr[$firstmultiletterword - 1] .=
"'";
1737 } elseif ( $firstspace > -1 ) {
1741 $arr[$firstspace] =
"''";
1742 $arr[$firstspace - 1] .=
"'";
1751 foreach ( $arr
as $r ) {
1752 if ( ( $i % 2 ) == 0 ) {
1753 if ( $state ===
'both' ) {
1759 $thislen = strlen( $r );
1760 if ( $thislen == 2 ) {
1761 if ( $state ===
'i' ) {
1764 } elseif ( $state ===
'bi' ) {
1767 } elseif ( $state ===
'ib' ) {
1770 } elseif ( $state ===
'both' ) {
1777 } elseif ( $thislen == 3 ) {
1778 if ( $state ===
'b' ) {
1781 } elseif ( $state ===
'bi' ) {
1784 } elseif ( $state ===
'ib' ) {
1787 } elseif ( $state ===
'both' ) {
1794 } elseif ( $thislen == 5 ) {
1795 if ( $state ===
'b' ) {
1798 } elseif ( $state ===
'i' ) {
1801 } elseif ( $state ===
'bi' ) {
1804 } elseif ( $state ===
'ib' ) {
1807 } elseif ( $state ===
'both' ) {
1819 if ( $state ===
'b' || $state ===
'ib' ) {
1822 if ( $state ===
'i' || $state ===
'bi' || $state ===
'ib' ) {
1825 if ( $state ===
'bi' ) {
1829 if ( $state ===
'both' &&
$buffer ) {
1848 public function replaceExternalLinks( $text ) {
1850 $bits = preg_split( $this->mExtLinkBracketedRegex, $text, -1, PREG_SPLIT_DELIM_CAPTURE );
1851 if ( $bits ===
false ) {
1852 throw new MWException(
"PCRE needs to be compiled with "
1853 .
"--enable-unicode-properties in order for MediaWiki to function" );
1855 $s = array_shift( $bits );
1858 while ( $i <
count( $bits ) ) {
1861 $text = $bits[$i++];
1862 $trail = $bits[$i++];
1864 # The characters '<' and '>' (which were escaped by
1865 # removeHTMLtags()) should not be included in
1866 # URLs, per RFC 2396.
1868 if ( preg_match(
'/&(lt|gt);/', $url, $m2, PREG_OFFSET_CAPTURE ) ) {
1869 $text = substr( $url, $m2[0][1] ) .
' ' . $text;
1870 $url = substr( $url, 0, $m2[0][1] );
1873 # If the link text is an image URL, replace it with an <img> tag
1874 # This happened by accident in the original parser, but some people used it extensively
1875 $img = $this->maybeMakeExternalImage( $text );
1876 if ( $img !==
false ) {
1882 # Set linktype for CSS - if URL==text, link is essentially free
1883 $linktype = ( $text === $url ) ?
'free' :
'text';
1885 # No link text, e.g. [http://domain.tld/some.link]
1886 if ( $text ==
'' ) {
1888 $langObj = $this->getTargetLanguage();
1889 $text =
'[' . $langObj->formatNum( ++$this->mAutonumber ) .
']';
1890 $linktype =
'autonumber';
1892 # Have link text, e.g. [http://domain.tld/some.link text]s
1897 $text = $this->getConverterLanguage()->markNoConversion( $text );
1899 $url = Sanitizer::cleanUrl( $url );
1901 # Use the encoded URL
1902 # This means that users can paste URLs directly into the text
1903 # Funny characters like ö aren't valid in URLs anyway
1904 # This was changed in August 2004
1906 $this->getExternalLinkAttribs( $url ), $this->mTitle ) . $dtrail . $trail;
1908 # Register link in the output object.
1909 $this->mOutput->addExternalLink( $url );
1924 public static function getExternalLinkRel( $url =
false,
$title =
null ) {
1925 global $wgNoFollowLinks, $wgNoFollowNsExceptions, $wgNoFollowDomainExceptions;
1927 if ( $wgNoFollowLinks && !in_array( $ns, $wgNoFollowNsExceptions )
1945 public function getExternalLinkAttribs( $url ) {
1947 $rel = self::getExternalLinkRel( $url, $this->mTitle );
1949 $target = $this->mOptions->getExternalLinkTarget();
1952 if ( !in_array( $target, [
'_self',
'_parent',
'_top' ] ) ) {
1956 if ( $rel !==
'' ) {
1959 $rel .=
'noreferrer noopener';
1975 public static function normalizeLinkUrl( $url ) {
1976 # First, make sure unsafe characters are encoded
1977 $url = preg_replace_callback(
'/[\x00-\x20"<>\[\\\\\]^`{|}\x7F-\xFF]/',
1979 return rawurlencode( $m[0] );
1985 $end = strlen( $url );
1987 # Fragment part - 'fragment'
1988 $start = strpos( $url,
'#' );
1989 if ( $start !==
false && $start < $end ) {
1990 $ret = self::normalizeUrlComponent(
1991 substr( $url, $start, $end - $start ),
'"#%<>[\]^`{|}' ) .
$ret;
1995 # Query part - 'query' minus &=+;
1996 $start = strpos( $url,
'?' );
1997 if ( $start !==
false && $start < $end ) {
1998 $ret = self::normalizeUrlComponent(
1999 substr( $url, $start, $end - $start ),
'"#%<>[\]^`{|}&=+;' ) .
$ret;
2003 # Scheme and path part - 'pchar'
2004 # (we assume no userinfo or encoded colons in the host)
2005 $ret = self::normalizeUrlComponent(
2006 substr( $url, 0, $end ),
'"#%<>[\]^`{|}/?' ) .
$ret;
2011 private static function normalizeUrlComponent( $component, $unsafe ) {
2012 $callback =
function (
$matches )
use ( $unsafe ) {
2014 $ord = ord( $char );
2015 if ( $ord > 32 && $ord < 127 && strpos( $unsafe, $char ) ===
false ) {
2019 # Leave it escaped, but use uppercase for a-f
2023 return preg_replace_callback(
'/%[0-9A-Fa-f]{2}/', $callback, $component );
2034 private function maybeMakeExternalImage( $url ) {
2035 $imagesfrom = $this->mOptions->getAllowExternalImagesFrom();
2036 $imagesexception = !empty( $imagesfrom );
2038 # $imagesfrom could be either a single string or an array of strings, parse out the latter
2039 if ( $imagesexception && is_array( $imagesfrom ) ) {
2040 $imagematch =
false;
2041 foreach ( $imagesfrom
as $match ) {
2042 if ( strpos( $url, $match ) === 0 ) {
2047 } elseif ( $imagesexception ) {
2048 $imagematch = ( strpos( $url, $imagesfrom ) === 0 );
2050 $imagematch =
false;
2053 if ( $this->mOptions->getAllowExternalImages()
2054 || ( $imagesexception && $imagematch )
2056 if ( preg_match( self::EXT_IMAGE_REGEX, $url ) ) {
2061 if ( !$text && $this->mOptions->getEnableImageWhitelist()
2062 && preg_match( self::EXT_IMAGE_REGEX, $url )
2064 $whitelist = explode(
2066 wfMessage(
'external_image_whitelist' )->inContentLanguage()->
text()
2069 foreach ( $whitelist
as $entry ) {
2070 # Sanitize the regex fragment, make it case-insensitive, ignore blank entries/comments
2071 if ( strpos( $entry,
'#' ) === 0 || $entry ===
'' ) {
2074 if ( preg_match(
'/' . str_replace(
'/',
'\\/', $entry ) .
'/i', $url ) ) {
2075 # Image matches a whitelist entry
2093 public function replaceInternalLinks(
$s ) {
2094 $this->mLinkHolders->merge( $this->replaceInternalLinks2(
$s ) );
2106 public function replaceInternalLinks2( &
$s ) {
2109 static $tc =
false, $e1, $e1_img;
2110 # the % is needed to support urlencoded titles as well
2113 # Match a link having the form [[namespace:link|alternate]]trail
2114 $e1 =
"/^([{$tc}]+)(?:\\|(.+?))?]](.*)\$/sD";
2115 # Match cases where there is no "]]", which might still be images
2116 $e1_img =
"/^([{$tc}]+)\\|(.*)\$/sD";
2121 # split the entire text string on occurrences of [[
2123 # get the first element (all text up to first [[), and remove the space we added
2126 $line = $a->current(); # Workaround
for broken ArrayIterator::next()
that returns "
void"
2127 $s = substr(
$s, 1 );
2129 $useLinkPrefixExtension = $this->getTargetLanguage()->linkPrefixExtension();
2131 if ( $useLinkPrefixExtension ) {
2132 # Match the end of a line for a word that's not followed by whitespace,
2133 # e.g. in the case of 'The Arab al[[Razi]]', 'al' will be matched
2136 $e2 =
"/^((?>.*[^$charset]|))(.+)$/sDu";
2139 if ( is_null( $this->mTitle ) ) {
2140 throw new MWException( __METHOD__ .
": \$this->mTitle is null\n" );
2142 $nottalk = !$this->mTitle->isTalkPage();
2144 if ( $useLinkPrefixExtension ) {
2146 if ( preg_match( $e2,
$s, $m ) ) {
2147 $first_prefix = $m[2];
2149 $first_prefix =
false;
2155 $useSubpages = $this->areSubpagesAllowed();
2158 # Loop for each link
2159 for ( ;
$line !==
false &&
$line !==
null; $a->next(),
$line = $a->current() ) {
2162 # Check for excessive memory usage
2163 if ( $holders->isBig() ) {
2165 # Do the existence check, replace the link holders and clear the array
2166 $holders->replace(
$s );
2170 if ( $useLinkPrefixExtension ) {
2171 if ( preg_match( $e2,
$s, $m ) ) {
2178 if ( $first_prefix ) {
2179 $prefix = $first_prefix;
2180 $first_prefix =
false;
2184 $might_be_img =
false;
2188 # If we get a ] at the beginning of $m[3] that means we have a link that's something like:
2189 # [[Image:Foo.jpg|[http://example.com desc]]] <- having three ] in a row fucks up,
2190 # the real problem is with the $e1 regex
2192 # Still some problems for cases where the ] is meant to be outside punctuation,
2193 # and no image is in sight. See T4095.
2195 && substr( $m[3], 0, 1 ) ===
']'
2196 && strpos( $text,
'[' ) !==
false
2198 $text .=
']'; #
so that replaceExternalLinks($text) works
later
2199 $m[3] = substr( $m[3], 1 );
2201 # fix up urlencoded title texts
2202 if ( strpos( $m[1],
'%' ) !==
false ) {
2203 # Should anchors '#' also be rejected?
2204 $m[1] = str_replace( [
'<',
'>' ], [
'<',
'>' ], rawurldecode( $m[1] ) );
2207 } elseif ( preg_match( $e1_img,
$line, $m ) ) {
2208 # Invalid, but might be an image with a link in its caption
2209 $might_be_img =
true;
2211 if ( strpos( $m[1],
'%' ) !==
false ) {
2212 $m[1] = str_replace( [
'<',
'>' ], [
'<',
'>' ], rawurldecode( $m[1] ) );
2220 $origLink = ltrim( $m[1],
' ' );
2222 # Don't allow internal links to pages containing
2223 # PROTO: where PROTO is a valid URL protocol; these
2224 # should be external links.
2225 if ( preg_match(
'/^(?i:' . $this->mUrlProtocols .
')/', $origLink ) ) {
2230 # Make subpage if necessary
2231 if ( $useSubpages ) {
2232 $link = $this->maybeDoSubpageLink( $origLink, $text );
2237 $noforce = ( substr( $origLink, 0, 1 ) !==
':' );
2239 # Strip off leading ':'
2243 $unstrip = $this->mStripState->unstripNoWiki(
$link );
2245 if ( $nt ===
null ) {
2250 $ns = $nt->getNamespace();
2251 $iw = $nt->getInterwiki();
2253 if ( $might_be_img ) { #
if this is actually an invalid
link
2254 if ( $ns ==
NS_FILE && $noforce ) { # but might be an image
2257 # look at the next 'line' to see if we can close it there
2259 $next_line = $a->current();
2260 if ( $next_line ===
false || $next_line ===
null ) {
2263 $m = explode(
']]', $next_line, 3 );
2264 if (
count( $m ) == 3 ) {
2265 # the first ]] closes the inner link, the second the image
2267 $text .=
"[[{$m[0]}]]{$m[1]}";
2270 } elseif (
count( $m ) == 2 ) {
2271 # if there's exactly one ]] that's fine, we'll keep looking
2272 $text .=
"[[{$m[0]}]]{$m[1]}";
2274 # if $next_line is invalid too, we need look no further
2275 $text .=
'[[' . $next_line;
2280 # we couldn't find the end of this imageLink, so output it raw
2281 # but don't ignore what might be perfectly normal links in the text we've examined
2282 $holders->merge( $this->replaceInternalLinks2( $text ) );
2283 $s .=
"{$prefix}[[$link|$text";
2284 # note: no $trail, because without an end, there *is* no trail
2287 }
else { #
it's not an image, so output it raw
2288 $s .= "{$prefix}[[$link|$text";
2289 # note: no $trail, because without an end, there *is* no trail
2294 $wasblank = ( $text == '' );
2298 # T6598 madness. Handle the quotes only if they come from the alternate part
2299 # [[Lista d''e paise d''o munno]] -> <a href="...">Lista d''e paise d''o munno</a>
2300 # [[Criticism of Harry Potter|Criticism of ''Harry Potter'']]
2301 # -> <a href="Criticism of Harry Potter">Criticism of <i>Harry Potter</i></a>
2302 $text = $this->doQuotes( $text );
2305 # Link not escaped by : , create the various objects
2306 if ( $noforce && !$nt->wasLocalInterwiki() ) {
2309 $iw && $this->mOptions->getInterwikiMagic() && $nottalk && (
2310 Language::fetchLanguageName( $iw, null, 'mw
' ) ||
2311 in_array( $iw, $wgExtraInterlanguageLinkPrefixes )
2314 # T26502: filter duplicates
2315 if ( !isset( $this->mLangLinkLanguages[$iw] ) ) {
2316 $this->mLangLinkLanguages[$iw] = true;
2317 $this->mOutput->addLanguageLink( $nt->getFullText() );
2320 $s = rtrim( $s . $prefix );
2321 $s .= trim( $trail, "\n" ) == '' ? '': $prefix . $trail;
2325 if ( $ns == NS_FILE ) {
2326 if ( !wfIsBadImage( $nt->getDBkey(), $this->mTitle ) ) {
2328 # if no parameters were passed, $text
2329 # becomes something like "File:Foo.png",
2330 # which we don't want to pass
on to the
2334 # recursively parse links inside the image caption
2335 # actually, this will parse them in any other parameters, too,
2336 # but it might be hard to fix that, and it doesn't matter ATM
2337 $text = $this->replaceExternalLinks( $text );
2338 $holders->merge( $this->replaceInternalLinks2( $text ) );
2340 # cloak any absolute URLs inside the image markup, so replaceExternalLinks() won't touch them
2341 $s .= $prefix . $this->armorLinks(
2342 $this->makeImage( $nt, $text, $holders ) ) . $trail;
2346 $s = rtrim(
$s .
"\n" ); # T2087
2349 $sortkey = $this->getDefaultSort();
2353 $sortkey = Sanitizer::decodeCharReferences( $sortkey );
2354 $sortkey = str_replace(
"\n",
'', $sortkey );
2355 $sortkey = $this->getConverterLanguage()->convertCategoryKey( $sortkey );
2356 $this->mOutput->addCategory( $nt->getDBkey(), $sortkey );
2361 $s .= trim( $prefix . $trail,
"\n" ) ==
'' ?
'' : $prefix . $trail;
2367 # Self-link checking. For some languages, variants of the title are checked in
2368 # LinkHolderArray::doVariants() to allow batching the existence checks necessary
2369 # for linking to a different variant.
2370 if ( $ns !=
NS_SPECIAL && $nt->equals( $this->mTitle ) && !$nt->hasFragment() ) {
2375 # NS_MEDIA is a pseudo-namespace for linking directly to a file
2376 # @todo FIXME: Should do batch file existence checks, see comment below
2378 # Give extensions a chance to select the file revision for us
2382 [ $this, $nt, &
$options, &$descQuery ] );
2383 # Fetch and register the file (file title may be different via hooks)
2384 list( $file, $nt ) = $this->fetchFileAndTitle( $nt,
$options );
2385 # Cloak with NOPARSE to avoid replacement in replaceExternalLinks
2386 $s .= $prefix . $this->armorLinks(
2391 # Some titles, such as valid special pages or files in foreign repos, should
2392 # be shown as bluelinks even though they're not included in the page table
2393 # @todo FIXME: isAlwaysKnown() can be expensive for file links; we should really do
2394 # batch file existence checks for NS_FILE and NS_MEDIA
2395 if ( $iw ==
'' && $nt->isAlwaysKnown() ) {
2396 $this->mOutput->addLink( $nt );
2397 $s .= $this->makeKnownLinkHolder( $nt, $text, $trail, $prefix );
2399 # Links will be added to the output link list after checking
2400 $s .= $holders->makeHolder( $nt, $text, [], $trail, $prefix );
2419 protected function makeKnownLinkHolder( $nt, $text =
'', $trail =
'', $prefix =
'' ) {
2422 if ( $text ==
'' ) {
2423 $text = htmlspecialchars( $nt->getPrefixedText() );
2426 $link = $this->getLinkRenderer()->makeKnownLink(
2427 $nt,
new HtmlArmor(
"$prefix$text$inside" )
2430 return $this->armorLinks(
$link ) . $trail;
2443 public function armorLinks( $text ) {
2444 return preg_replace(
'/\b((?i)' . $this->mUrlProtocols .
')/',
2445 self::MARKER_PREFIX .
"NOPARSE$1", $text );
2452 public function areSubpagesAllowed() {
2453 # Some namespaces don't allow subpages
2465 public function maybeDoSubpageLink( $target, &$text ) {
2477 public function doBlockLevels( $text, $linestart ) {
2492 public function getVariableValue( $index, $frame =
false ) {
2496 if ( is_null( $this->mTitle ) ) {
2501 throw new MWException( __METHOD__ .
' Should only be '
2502 .
' called while parsing (no title set)' );
2512 if (
Hooks::run(
'ParserGetVariableValueVarCache', [ &
$parser, &$this->mVarCache ] ) ) {
2513 if ( isset( $this->mVarCache[$index] ) ) {
2514 return $this->mVarCache[$index];
2518 $ts =
wfTimestamp( TS_UNIX, $this->mOptions->getTimestamp() );
2521 $pageLang = $this->getFunctionLang();
2527 case 'currentmonth':
2530 case 'currentmonth1':
2533 case 'currentmonthname':
2536 case 'currentmonthnamegen':
2539 case 'currentmonthabbrev':
2554 case 'localmonthname':
2557 case 'localmonthnamegen':
2560 case 'localmonthabbrev':
2575 case 'fullpagename':
2578 case 'fullpagenamee':
2584 case 'subpagenamee':
2587 case 'rootpagename':
2590 case 'rootpagenamee':
2594 $this->mTitle->getRootText()
2597 case 'basepagename':
2600 case 'basepagenamee':
2604 $this->mTitle->getBaseText()
2607 case 'talkpagename':
2608 if ( $this->mTitle->canTalk() ) {
2609 $talkPage = $this->mTitle->getTalkPage();
2615 case 'talkpagenamee':
2616 if ( $this->mTitle->canTalk() ) {
2617 $talkPage = $this->mTitle->getTalkPage();
2623 case 'subjectpagename':
2624 $subjPage = $this->mTitle->getSubjectPage();
2627 case 'subjectpagenamee':
2628 $subjPage = $this->mTitle->getSubjectPage();
2632 $pageid = $this->getTitle()->getArticleID();
2633 if ( $pageid == 0 ) {
2634 # 0 means the page doesn't exist in the database,
2635 # which means the user is previewing a new page.
2636 # The vary-revision flag must be set, because the magic word
2637 # will have a different value once the page is saved.
2638 $this->mOutput->setFlag(
'vary-revision' );
2639 wfDebug( __METHOD__ .
": {{PAGEID}} used in a new page, setting vary-revision...\n" );
2641 $value = $pageid ? $pageid :
null;
2644 # Let the edit saving system know we should parse the page
2645 # *after* a revision ID has been assigned.
2646 $this->mOutput->setFlag(
'vary-revision-id' );
2647 wfDebug( __METHOD__ .
": {{REVISIONID}} used, setting vary-revision-id...\n" );
2648 $value = $this->mRevisionId;
2649 if ( !
$value && $this->mOptions->getSpeculativeRevIdCallback() ) {
2650 $value = call_user_func( $this->mOptions->getSpeculativeRevIdCallback() );
2651 $this->mOutput->setSpeculativeRevIdUsed(
$value );
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__ .
": {{REVISIONDAY}} used, setting vary-revision...\n" );
2659 $value = intval( substr( $this->getRevisionTimestamp(), 6, 2 ) );
2661 case 'revisionday2':
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__ .
": {{REVISIONDAY2}} used, setting vary-revision...\n" );
2666 $value = substr( $this->getRevisionTimestamp(), 6, 2 );
2668 case 'revisionmonth':
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__ .
": {{REVISIONMONTH}} used, setting vary-revision...\n" );
2673 $value = substr( $this->getRevisionTimestamp(), 4, 2 );
2675 case 'revisionmonth1':
2676 # Let the edit saving system know we should parse the page
2677 # *after* a revision ID has been assigned. This is for null edits.
2678 $this->mOutput->setFlag(
'vary-revision' );
2679 wfDebug( __METHOD__ .
": {{REVISIONMONTH1}} used, setting vary-revision...\n" );
2680 $value = intval( substr( $this->getRevisionTimestamp(), 4, 2 ) );
2682 case 'revisionyear':
2683 # Let the edit saving system know we should parse the page
2684 # *after* a revision ID has been assigned. This is for null edits.
2685 $this->mOutput->setFlag(
'vary-revision' );
2686 wfDebug( __METHOD__ .
": {{REVISIONYEAR}} used, setting vary-revision...\n" );
2687 $value = substr( $this->getRevisionTimestamp(), 0, 4 );
2689 case 'revisiontimestamp':
2690 # Let the edit saving system know we should parse the page
2691 # *after* a revision ID has been assigned. This is for null edits.
2692 $this->mOutput->setFlag(
'vary-revision' );
2693 wfDebug( __METHOD__ .
": {{REVISIONTIMESTAMP}} used, setting vary-revision...\n" );
2694 $value = $this->getRevisionTimestamp();
2696 case 'revisionuser':
2697 # Let the edit saving system know we should parse the page
2698 # *after* a revision ID has been assigned for null edits.
2699 $this->mOutput->setFlag(
'vary-user' );
2700 wfDebug( __METHOD__ .
": {{REVISIONUSER}} used, setting vary-user...\n" );
2701 $value = $this->getRevisionUser();
2703 case 'revisionsize':
2704 $value = $this->getRevisionSize();
2707 $value = str_replace(
'_',
' ',
$wgContLang->getNsText( $this->mTitle->getNamespace() ) );
2712 case 'namespacenumber':
2713 $value = $this->mTitle->getNamespace();
2716 $value = $this->mTitle->canTalk()
2717 ? str_replace(
'_',
' ', $this->mTitle->getTalkNsText() )
2721 $value = $this->mTitle->canTalk() ?
wfUrlencode( $this->mTitle->getTalkNsText() ) :
'';
2723 case 'subjectspace':
2724 $value = str_replace(
'_',
' ', $this->mTitle->getSubjectNsText() );
2726 case 'subjectspacee':
2729 case 'currentdayname':
2742 # @bug T6594 PHP5 has it zero padded, PHP4 does not, cast to
2743 # int to remove the padding
2749 case 'localdayname':
2750 $value = $pageLang->getWeekdayName(
2758 $value = $pageLang->time(
2768 # @bug T6594 PHP5 has it zero padded, PHP4 does not, cast to
2769 # int to remove the padding
2775 case 'numberofarticles':
2778 case 'numberoffiles':
2781 case 'numberofusers':
2784 case 'numberofactiveusers':
2787 case 'numberofpages':
2790 case 'numberofadmins':
2793 case 'numberofedits':
2796 case 'currenttimestamp':
2799 case 'localtimestamp':
2802 case 'currentversion':
2817 case 'directionmark':
2818 return $pageLang->getDirMark();
2819 case 'contentlanguage':
2822 case 'pagelanguage':
2823 $value = $pageLang->getCode();
2825 case 'cascadingsources':
2831 'ParserGetVariableValueSwitch',
2832 [ &
$parser, &$this->mVarCache, &$index, &
$ret, &$frame ]
2839 $this->mVarCache[$index] =
$value;
2850 public function initialiseVariables() {
2880 public function preprocessToDom( $text,
$flags = 0 ) {
2881 $dom = $this->getPreprocessor()->preprocessToObj( $text,
$flags );
2892 public static function splitWhitespace(
$s ) {
2893 $ltrimmed = ltrim(
$s );
2894 $w1 = substr(
$s, 0, strlen(
$s ) - strlen( $ltrimmed ) );
2895 $trimmed = rtrim( $ltrimmed );
2896 $diff = strlen( $ltrimmed ) - strlen( $trimmed );
2898 $w2 = substr( $ltrimmed, -$diff );
2902 return [ $w1, $trimmed, $w2 ];
2925 public function replaceVariables( $text, $frame =
false, $argsOnly =
false ) {
2926 # Is there any text? Also, Prevent too big inclusions!
2927 $textSize = strlen( $text );
2928 if ( $textSize < 1 || $textSize > $this->mOptions->getMaxIncludeSize() ) {
2932 if ( $frame ===
false ) {
2933 $frame = $this->getPreprocessor()->newFrame();
2934 } elseif ( !( $frame instanceof
PPFrame ) ) {
2935 wfDebug( __METHOD__ .
" called using plain parameters instead of "
2936 .
"a PPFrame instance. Creating custom frame.\n" );
2937 $frame = $this->getPreprocessor()->newCustomFrame( $frame );
2940 $dom = $this->preprocessToDom( $text );
2942 $text = $frame->expand( $dom,
$flags );
2954 public static function createAssocArgs(
$args ) {
2958 $eqpos = strpos( $arg,
'=' );
2959 if ( $eqpos ===
false ) {
2960 $assocArgs[$index++] = $arg;
2962 $name = trim( substr( $arg, 0, $eqpos ) );
2963 $value = trim( substr( $arg, $eqpos + 1 ) );
2964 if (
$value ===
false ) {
2967 if (
$name !==
false ) {
3002 public function limitationWarn( $limitationType, $current =
'', $max =
'' ) {
3003 # does no harm if $current and $max are present but are unnecessary for the message
3004 # Not doing ->inLanguage( $this->mOptions->getUserLangObj() ), since this is shown
3005 # only during preview, and that would split the parser cache unnecessarily.
3006 $warning =
wfMessage(
"$limitationType-warning" )->numParams( $current, $max )
3008 $this->mOutput->addWarning( $warning );
3009 $this->addTrackingCategory(
"$limitationType-category" );
3024 public function braceSubstitution( $piece, $frame ) {
3035 $forceRawInterwiki =
false;
3037 $isChildObj =
false;
3039 $isLocalObj =
false;
3041 # Title object, where $text came from
3044 # $part1 is the bit before the first |, and must contain only title characters.
3045 # Various prefixes will be stripped from it later.
3046 $titleWithSpaces = $frame->expand( $piece[
'title'] );
3047 $part1 = trim( $titleWithSpaces );
3050 # Original title text preserved for various purposes
3051 $originalTitle = $part1;
3053 # $args is a list of argument nodes, starting from index 0, not including $part1
3054 # @todo FIXME: If piece['parts'] is null then the call to getLength()
3055 # below won't work b/c this $args isn't an object
3056 $args = (
null == $piece[
'parts'] ) ? [] : $piece[
'parts'];
3058 $profileSection =
null;
3062 $substMatch = $this->mSubstWords->matchStartAndRemove( $part1 );
3064 # Possibilities for substMatch: "subst", "safesubst" or FALSE
3065 # Decide whether to expand template or keep wikitext as-is.
3066 if ( $this->ot[
'wiki'] ) {
3067 if ( $substMatch ===
false ) {
3068 $literal =
true; # literal when
in PST with no prefix
3070 $literal =
false; # expand when
in PST with subst:
or safesubst:
3073 if ( $substMatch ==
'subst' ) {
3074 $literal =
true; # literal when
not in PST with
plain subst:
3076 $literal =
false; # expand when
not in PST with safesubst:
or no prefix
3080 $text = $frame->virtualBracketedImplode(
'{{',
'|',
'}}', $titleWithSpaces,
$args );
3087 if ( !$found &&
$args->getLength() == 0 ) {
3088 $id = $this->mVariables->matchStartToEnd( $part1 );
3089 if ( $id !==
false ) {
3090 $text = $this->getVariableValue( $id, $frame );
3098 # MSG, MSGNW and RAW
3102 if ( $mwMsgnw->matchStartAndRemove( $part1 ) ) {
3105 # Remove obsolete MSG:
3107 $mwMsg->matchStartAndRemove( $part1 );
3112 if ( $mwRaw->matchStartAndRemove( $part1 ) ) {
3113 $forceRawInterwiki =
true;
3119 $colonPos = strpos( $part1,
':' );
3120 if ( $colonPos !==
false ) {
3121 $func = substr( $part1, 0, $colonPos );
3122 $funcArgs = [ trim( substr( $part1, $colonPos + 1 ) ) ];
3123 $argsLength =
$args->getLength();
3124 for ( $i = 0; $i < $argsLength; $i++ ) {
3125 $funcArgs[] =
$args->item( $i );
3128 $result = $this->callParserFunction( $frame, $func, $funcArgs );
3129 }
catch ( Exception $ex ) {
3133 # The interface for parser functions allows for extracting
3134 # flags into the local scope. Extract any forwarded flags
3140 # Finish mangling title and then check for loops.
3141 # Set $title to a Title object and $titleText to the PDBK
3144 # Split the title into page and subpage
3146 $relative = $this->maybeDoSubpageLink( $part1, $subpage );
3147 if ( $part1 !== $relative ) {
3149 $ns = $this->mTitle->getNamespace();
3153 $titleText =
$title->getPrefixedText();
3154 # Check for language variants if the template is not found
3155 if ( $this->getConverterLanguage()->hasVariants() &&
$title->getArticleID() == 0 ) {
3156 $this->getConverterLanguage()->findVariantLink( $part1,
$title,
true );
3158 # Do recursion depth check
3159 $limit = $this->mOptions->getMaxTemplateDepth();
3160 if ( $frame->depth >=
$limit ) {
3162 $text =
'<span class="error">'
3163 .
wfMessage(
'parser-template-recursion-depth-warning' )
3164 ->numParams(
$limit )->inContentLanguage()->text()
3170 # Load from database
3171 if ( !$found &&
$title ) {
3172 $profileSection = $this->mProfiler->scopedProfileIn(
$title->getPrefixedDBkey() );
3173 if ( !
$title->isExternal() ) {
3174 if (
$title->isSpecialPage()
3175 && $this->mOptions->getAllowSpecialInclusion()
3176 && $this->ot[
'html']
3183 $argsLength =
$args->getLength();
3184 for ( $i = 0; $i < $argsLength; $i++ ) {
3185 $bits =
$args->item( $i )->splitArg();
3186 if ( strval( $bits[
'index'] ) ===
'' ) {
3188 $value = trim( $frame->expand( $bits[
'value'] ) );
3197 if ( $specialPage && $specialPage->maxIncludeCacheTime() === 0 ) {
3198 $context->setUser( $this->getUser() );
3203 $context->setLanguage( $this->mOptions->getUserLangObj() );
3211 if ( $specialPage && $specialPage->maxIncludeCacheTime() !==
false ) {
3212 $this->mOutput->updateRuntimeAdaptiveExpiry(
3213 $specialPage->maxIncludeCacheTime()
3218 $found =
false; # access denied
3219 wfDebug( __METHOD__ .
": template inclusion denied for " .
3220 $title->getPrefixedDBkey() .
"\n" );
3223 if ( $text !==
false ) {
3229 # If the title is valid but undisplayable, make a link to it
3230 if ( !$found && ( $this->ot[
'html'] || $this->ot[
'pre'] ) ) {
3231 $text =
"[[:$titleText]]";
3234 } elseif (
$title->isTrans() ) {
3235 # Interwiki transclusion
3236 if ( $this->ot[
'html'] && !$forceRawInterwiki ) {
3237 $text = $this->interwikiTransclude(
$title,
'render' );
3240 $text = $this->interwikiTransclude(
$title,
'raw' );
3241 # Preprocess it like a template
3242 $text = $this->preprocessToDom( $text, self::PTD_FOR_INCLUSION );
3248 # Do infinite loop check
3249 # This has to be done after redirect resolution to avoid infinite loops via redirects
3250 if ( !$frame->loopCheck(
$title ) ) {
3252 $text =
'<span class="error">'
3253 .
wfMessage(
'parser-template-loop-warning', $titleText )->inContentLanguage()->text()
3255 $this->addTrackingCategory(
'template-loop-category' );
3256 wfDebug( __METHOD__ .
": template loop broken at '$titleText'\n" );
3260 # If we haven't found text to substitute by now, we're done
3261 # Recover the source wikitext and return it
3263 $text = $frame->virtualBracketedImplode(
'{{',
'|',
'}}', $titleWithSpaces,
$args );
3264 if ( $profileSection ) {
3265 $this->mProfiler->scopedProfileOut( $profileSection );
3267 return [
'object' => $text ];
3270 # Expand DOM-style return values in a child frame
3271 if ( $isChildObj ) {
3272 # Clean up argument array
3277 } elseif ( $titleText !==
false && $newFrame->isEmpty() ) {
3278 # Expansion is eligible for the empty-frame cache
3279 $text = $newFrame->cachedExpand( $titleText, $text );
3281 # Uncached expansion
3282 $text = $newFrame->expand( $text );
3285 if ( $isLocalObj && $nowiki ) {
3287 $isLocalObj =
false;
3290 if ( $profileSection ) {
3291 $this->mProfiler->scopedProfileOut( $profileSection );
3294 # Replace raw HTML by a placeholder
3296 $text = $this->insertStripItem( $text );
3297 } elseif ( $nowiki && ( $this->ot[
'html'] || $this->ot[
'pre'] ) ) {
3298 # Escape nowiki-style return values
3300 } elseif ( is_string( $text )
3301 && !$piece[
'lineStart']
3302 && preg_match(
'/^(?:{\\||:|;|#|\*)/', $text )
3304 # T2529: if the template begins with a table or block-level
3305 # element, it should be treated as beginning a new line.
3306 # This behavior is somewhat controversial.
3307 $text =
"\n" . $text;
3310 if ( is_string( $text ) && !$this->incrementIncludeSize(
'post-expand', strlen( $text ) ) ) {
3311 # Error, oversize inclusion
3312 if ( $titleText !==
false ) {
3313 # Make a working, properly escaped link if possible (T25588)
3314 $text =
"[[:$titleText]]";
3316 # This will probably not be a working link, but at least it may
3317 # provide some hint of where the problem is
3318 preg_replace(
'/^:/',
'', $originalTitle );
3319 $text =
"[[:$originalTitle]]";
3321 $text .= $this->insertStripItem(
'<!-- WARNING: template omitted, '
3322 .
'post-expand include size too large -->' );
3323 $this->limitationWarn(
'post-expand-template-inclusion' );
3326 if ( $isLocalObj ) {
3327 $ret = [
'object' => $text ];
3329 $ret = [
'text' => $text ];
3354 public function callParserFunction( $frame, $function,
array $args = [] ) {
3357 # Case sensitive functions
3358 if ( isset( $this->mFunctionSynonyms[1][$function] ) ) {
3359 $function = $this->mFunctionSynonyms[1][$function];
3361 # Case insensitive functions
3363 if ( isset( $this->mFunctionSynonyms[0][$function] ) ) {
3364 $function = $this->mFunctionSynonyms[0][$function];
3366 return [
'found' =>
false ];
3370 list( $callback,
$flags ) = $this->mFunctionHooks[$function];
3372 # Workaround for PHP bug 35229 and similar
3373 if ( !is_callable( $callback ) ) {
3374 throw new MWException(
"Tag hook for $function is not callable\n" );
3382 # Convert arguments to PPNodes and collect for appending to $allArgs
3384 foreach (
$args as $k => $v ) {
3385 if ( $v instanceof
PPNode || $k === 0 ) {
3388 $funcArgs[] = $this->mPreprocessor->newPartNodeArray( [ $k => $v ] )->item( 0 );
3392 # Add a frame parameter, and pass the arguments as an array
3393 $allArgs[] = $frame;
3394 $allArgs[] = $funcArgs;
3396 # Convert arguments to plain text and append to $allArgs
3397 foreach (
$args as $k => $v ) {
3398 if ( $v instanceof
PPNode ) {
3399 $allArgs[] = trim( $frame->expand( $v ) );
3400 } elseif ( is_int( $k ) && $k >= 0 ) {
3401 $allArgs[] = trim( $v );
3403 $allArgs[] = trim(
"$k=$v" );
3408 $result = call_user_func_array( $callback, $allArgs );
3410 # The interface for function hooks allows them to return a wikitext
3411 # string or an array containing the string and any flags. This mungs
3412 # things around to match what this method should return.
3429 $preprocessFlags = 0;
3430 if ( isset(
$result[
'noparse'] ) ) {
3431 $noparse =
$result[
'noparse'];
3433 if ( isset(
$result[
'preprocessFlags'] ) ) {
3434 $preprocessFlags =
$result[
'preprocessFlags'];
3438 $result[
'text'] = $this->preprocessToDom(
$result[
'text'], $preprocessFlags );
3453 public function getTemplateDom(
$title ) {
3455 $titleText =
$title->getPrefixedDBkey();
3457 if ( isset( $this->mTplRedirCache[$titleText] ) ) {
3458 list( $ns, $dbk ) = $this->mTplRedirCache[$titleText];
3460 $titleText =
$title->getPrefixedDBkey();
3462 if ( isset( $this->mTplDomCache[$titleText] ) ) {
3463 return [ $this->mTplDomCache[$titleText],
$title ];
3466 # Cache miss, go to the database
3469 if ( $text ===
false ) {
3470 $this->mTplDomCache[$titleText] =
false;
3471 return [
false,
$title ];
3474 $dom = $this->preprocessToDom( $text, self::PTD_FOR_INCLUSION );
3475 $this->mTplDomCache[$titleText] = $dom;
3477 if ( !
$title->equals( $cacheTitle ) ) {
3478 $this->mTplRedirCache[$cacheTitle->getPrefixedDBkey()] =
3496 public function fetchCurrentRevisionOfTitle(
$title ) {
3497 $cacheKey =
$title->getPrefixedDBkey();
3498 if ( !$this->currentRevisionCache ) {
3499 $this->currentRevisionCache =
new MapCacheLRU( 100 );
3501 if ( !$this->currentRevisionCache->has( $cacheKey ) ) {
3502 $this->currentRevisionCache->set( $cacheKey,
3504 call_user_func( $this->mOptions->getCurrentRevisionCallback(),
$title, $this )
3507 return $this->currentRevisionCache->get( $cacheKey );
3520 $pageId =
$title->getArticleID();
3536 public function fetchTemplateAndTitle(
$title ) {
3538 $templateCb = $this->mOptions->getTemplateCallback();
3539 $stuff = call_user_func( $templateCb,
$title, $this );
3541 $text = $stuff[
'text'];
3542 if ( is_string( $stuff[
'text'] ) ) {
3543 $text = strtr( $text,
"\x7f",
"?" );
3545 $finalTitle = isset( $stuff[
'finalTitle'] ) ? $stuff[
'finalTitle'] :
$title;
3546 if ( isset( $stuff[
'deps'] ) ) {
3547 foreach ( $stuff[
'deps']
as $dep ) {
3548 $this->mOutput->addTemplate( $dep[
'title'], $dep[
'page_id'], $dep[
'rev_id'] );
3549 if ( $dep[
'title']->equals( $this->getTitle() ) ) {
3552 $this->mOutput->setFlag(
'vary-revision' );
3556 return [ $text, $finalTitle ];
3564 public function fetchTemplate(
$title ) {
3565 return $this->fetchTemplateAndTitle(
$title )[0];
3577 public static function statelessFetchTemplate(
$title,
$parser =
false ) {
3578 $text = $skip =
false;
3582 # Loop to fetch the article, with up to 1 redirect
3584 for ( $i = 0; $i < 2 && is_object(
$title ); $i++ ) {
3586 # Give extensions a chance to select the revision instead
3587 $id =
false; # Assume current
3588 Hooks::run(
'BeforeParserFetchTemplateAndtitle',
3595 'page_id' =>
$title->getArticleID(),
3608 $rev_id =
$rev ?
$rev->getId() : 0;
3609 # If there is no current revision, there is no page
3610 if ( $id ===
false && !
$rev ) {
3612 $linkCache->addBadLinkObj(
$title );
3617 'page_id' =>
$title->getArticleID(),
3618 'rev_id' => $rev_id ];
3620 # We fetched a rev from a different title; register it too...
3622 'title' =>
$rev->getTitle(),
3623 'page_id' =>
$rev->getPage(),
3624 'rev_id' => $rev_id ];
3634 if ( $text ===
false || $text ===
null ) {
3641 if ( !$message->exists() ) {
3646 $text = $message->plain();
3659 'finalTitle' => $finalTitle,
3684 $time = $file ? $file->getTimestamp() :
false;
3685 $sha1 = $file ? $file->getSha1() :
false;
3686 # Register the file as a dependency...
3687 $this->mOutput->addImage(
$title->getDBkey(),
$time, $sha1 );
3688 if ( $file && !
$title->equals( $file->getTitle() ) ) {
3689 # Update fetched file title
3690 $title = $file->getTitle();
3691 $this->mOutput->addImage(
$title->getDBkey(),
$time, $sha1 );
3693 return [ $file,
$title ];
3706 protected function fetchFileNoRegister(
$title,
$options = [] ) {
3707 if ( isset(
$options[
'broken'] ) ) {
3709 } elseif ( isset(
$options[
'sha1'] ) ) {
3725 public function interwikiTransclude(
$title, $action ) {
3726 global $wgEnableScaryTranscluding;
3728 if ( !$wgEnableScaryTranscluding ) {
3729 return wfMessage(
'scarytranscludedisabled' )->inContentLanguage()->text();
3732 $url =
$title->getFullURL( [
'action' => $action ] );
3734 if ( strlen( $url ) > 255 ) {
3735 return wfMessage(
'scarytranscludetoolong' )->inContentLanguage()->text();
3737 return $this->fetchScaryTemplateMaybeFromCache( $url );
3744 public function fetchScaryTemplateMaybeFromCache( $url ) {
3745 global $wgTranscludeCacheExpiry;
3747 $tsCond =
$dbr->timestamp( time() - $wgTranscludeCacheExpiry );
3748 $obj =
$dbr->selectRow(
'transcache', [
'tc_time',
'tc_contents' ],
3749 [
'tc_url' => $url,
"tc_time >= " .
$dbr->addQuotes( $tsCond ) ] );
3751 return $obj->tc_contents;
3757 $text =
$req->getContent();
3758 } elseif (
$req->getStatus() != 200 ) {
3760 return wfMessage(
'scarytranscludefailed-httpstatus' )
3761 ->params( $url,
$req->getStatus() )->inContentLanguage()->text();
3763 return wfMessage(
'scarytranscludefailed', $url )->inContentLanguage()->text();
3767 $dbw->replace(
'transcache', [
'tc_url' ], [
3769 'tc_time' => $dbw->timestamp( time() ),
3770 'tc_contents' => $text
3784 public function argSubstitution( $piece, $frame ) {
3787 $parts = $piece[
'parts'];
3788 $nameWithSpaces = $frame->expand( $piece[
'title'] );
3789 $argName = trim( $nameWithSpaces );
3791 $text = $frame->getArgument( $argName );
3792 if ( $text ===
false && $parts->getLength() > 0
3793 && ( $this->ot[
'html']
3795 || ( $this->ot[
'wiki'] && $frame->isTemplate() )
3798 # No match in frame, use the supplied default
3799 $object = $parts->item( 0 )->getChildren();
3801 if ( !$this->incrementIncludeSize(
'arg', strlen( $text ) ) ) {
3802 $error =
'<!-- WARNING: argument omitted, expansion size too large -->';
3803 $this->limitationWarn(
'post-expand-template-argument' );
3806 if ( $text ===
false && $object ===
false ) {
3808 $object = $frame->virtualBracketedImplode(
'{{{',
'|',
'}}}', $nameWithSpaces, $parts );
3810 if ( $error !==
false ) {
3813 if ( $object !==
false ) {
3814 $ret = [
'object' => $object ];
3816 $ret = [
'text' => $text ];
3837 public function extensionSubstitution(
$params, $frame ) {
3838 static $errorStr =
'<span class="error">';
3839 static $errorLen = 20;
3842 if ( substr(
$name, 0, $errorLen ) === $errorStr ) {
3848 $attrText = !isset(
$params[
'attr'] ) ? null : $frame->expand(
$params[
'attr'] );
3849 if ( substr( $attrText, 0, $errorLen ) === $errorStr ) {
3859 $marker = self::MARKER_PREFIX .
"-$name-"
3860 . sprintf(
'%08X', $this->mMarkerIndex++ ) . self::MARKER_SUFFIX;
3862 $isFunctionTag = isset( $this->mFunctionTagHooks[strtolower(
$name )] ) &&
3863 ( $this->ot[
'html'] || $this->ot[
'pre'] );
3864 if ( $isFunctionTag ) {
3865 $markerType =
'none';
3867 $markerType =
'general';
3869 if ( $this->ot[
'html'] || $isFunctionTag ) {
3871 $attributes = Sanitizer::decodeTagAttributes( $attrText );
3872 if ( isset(
$params[
'attributes'] ) ) {
3873 $attributes = $attributes +
$params[
'attributes'];
3876 if ( isset( $this->mTagHooks[
$name] ) ) {
3877 # Workaround for PHP bug 35229 and similar
3878 if ( !is_callable( $this->mTagHooks[
$name] ) ) {
3879 throw new MWException(
"Tag hook for $name is not callable\n" );
3881 $output = call_user_func_array( $this->mTagHooks[
$name],
3882 [
$content, $attributes, $this, $frame ] );
3883 } elseif ( isset( $this->mFunctionTagHooks[
$name] ) ) {
3884 list( $callback, ) = $this->mFunctionTagHooks[
$name];
3885 if ( !is_callable( $callback ) ) {
3886 throw new MWException(
"Tag hook for $name is not callable\n" );
3893 $output =
'<span class="error">Invalid tag extension name: ' .
3894 htmlspecialchars(
$name ) .
'</span>';
3898 # Extract flags to local scope (to override $markerType)
3905 if ( is_null( $attrText ) ) {
3908 if ( isset(
$params[
'attributes'] ) ) {
3909 foreach (
$params[
'attributes']
as $attrName => $attrValue ) {
3910 $attrText .=
' ' . htmlspecialchars( $attrName ) .
'="' .
3911 htmlspecialchars( $attrValue ) .
'"';
3915 $output =
"<$name$attrText/>";
3917 $close = is_null(
$params[
'close'] ) ?
'' : $frame->expand(
$params[
'close'] );
3918 if ( substr( $close, 0, $errorLen ) === $errorStr ) {
3922 $output =
"<$name$attrText>$content$close";
3926 if ( $markerType ===
'none' ) {
3928 } elseif ( $markerType ===
'nowiki' ) {
3929 $this->mStripState->addNoWiki( $marker,
$output );
3930 } elseif ( $markerType ===
'general' ) {
3931 $this->mStripState->addGeneral( $marker,
$output );
3933 throw new MWException( __METHOD__ .
': invalid marker type' );
3945 public function incrementIncludeSize(
$type, $size ) {
3946 if ( $this->mIncludeSizes[
$type] + $size > $this->mOptions->getMaxIncludeSize() ) {
3949 $this->mIncludeSizes[
$type] += $size;
3959 public function incrementExpensiveFunctionCount() {
3960 $this->mExpensiveFunctionCount++;
3961 return $this->mExpensiveFunctionCount <= $this->mOptions->getExpensiveParserFunctionLimit();
3972 public function doDoubleUnderscore( $text ) {
3974 # The position of __TOC__ needs to be recorded
3976 if ( $mw->match( $text ) ) {
3977 $this->mShowToc =
true;
3978 $this->mForceTocPosition =
true;
3980 # Set a placeholder. At the end we'll fill it in with the TOC.
3981 $text = $mw->replace(
'<!--MWTOC-->', $text, 1 );
3983 # Only keep the first one.
3984 $text = $mw->replace(
'', $text );
3987 # Now match and remove the rest of them
3989 $this->mDoubleUnderscores = $mwa->matchAndRemove( $text );
3991 if ( isset( $this->mDoubleUnderscores[
'nogallery'] ) ) {
3992 $this->mOutput->mNoGallery =
true;
3994 if ( isset( $this->mDoubleUnderscores[
'notoc'] ) && !$this->mForceTocPosition ) {
3995 $this->mShowToc =
false;
3997 if ( isset( $this->mDoubleUnderscores[
'hiddencat'] )
4000 $this->addTrackingCategory(
'hidden-category-category' );
4002 # (T10068) Allow control over whether robots index a page.
4003 # __INDEX__ always overrides __NOINDEX__, see T16899
4004 if ( isset( $this->mDoubleUnderscores[
'noindex'] ) && $this->mTitle->canUseNoindex() ) {
4005 $this->mOutput->setIndexPolicy(
'noindex' );
4006 $this->addTrackingCategory(
'noindex-category' );
4008 if ( isset( $this->mDoubleUnderscores[
'index'] ) && $this->mTitle->canUseNoindex() ) {
4009 $this->mOutput->setIndexPolicy(
'index' );
4010 $this->addTrackingCategory(
'index-category' );
4013 # Cache all double underscores in the database
4014 foreach ( $this->mDoubleUnderscores
as $key => $val ) {
4015 $this->mOutput->setProperty( $key,
'' );
4026 public function addTrackingCategory( $msg ) {
4027 return $this->mOutput->addTrackingCategory( $msg, $this->mTitle );
4046 public function formatHeadings( $text, $origText, $isMain =
true ) {
4047 global $wgMaxTocLevel, $wgExperimentalHtmlIds;
4049 # Inhibit editsection links if requested in the page
4050 if ( isset( $this->mDoubleUnderscores[
'noeditsection'] ) ) {
4051 $maybeShowEditLink = $showEditLink =
false;
4053 $maybeShowEditLink =
true;
4054 $showEditLink = $this->mOptions->getEditSection();
4056 if ( $showEditLink ) {
4057 $this->mOutput->setEditSectionTokens(
true );
4060 # Get all headlines for numbering them and adding funky stuff like [edit]
4061 # links - this is for later, but we need the number of headlines right now
4063 $numMatches = preg_match_all(
4064 '/<H(?P<level>[1-6])(?P<attrib>.*?>)\s*(?P<header>[\s\S]*?)\s*<\/H[1-6] *>/i',
4069 # if there are fewer than 4 headlines in the article, do not show TOC
4070 # unless it's been explicitly enabled.
4071 $enoughToc = $this->mShowToc &&
4072 ( ( $numMatches >= 4 ) || $this->mForceTocPosition );
4074 # Allow user to stipulate that a page should have a "new section"
4075 # link added via __NEWSECTIONLINK__
4076 if ( isset( $this->mDoubleUnderscores[
'newsectionlink'] ) ) {
4077 $this->mOutput->setNewSection(
true );
4080 # Allow user to remove the "new section"
4081 # link via __NONEWSECTIONLINK__
4082 if ( isset( $this->mDoubleUnderscores[
'nonewsectionlink'] ) ) {
4083 $this->mOutput->hideNewSection(
true );
4086 # if the string __FORCETOC__ (not case-sensitive) occurs in the HTML,
4087 # override above conditions and always show TOC above first header
4088 if ( isset( $this->mDoubleUnderscores[
'forcetoc'] ) ) {
4089 $this->mShowToc =
true;
4097 # Ugh .. the TOC should have neat indentation levels which can be
4098 # passed to the skin functions. These are determined here
4102 $sublevelCount = [];
4108 $markerRegex = self::MARKER_PREFIX .
"-h-(\d+)-" . self::MARKER_SUFFIX;
4109 $baseTitleText = $this->mTitle->getPrefixedDBkey();
4110 $oldType = $this->mOutputType;
4112 $frame = $this->getPreprocessor()->newFrame();
4113 $root = $this->preprocessToDom( $origText );
4114 $node = $root->getFirstChild();
4119 $headlines = $numMatches !==
false ?
$matches[3] : [];
4121 foreach ( $headlines
as $headline ) {
4122 $isTemplate =
false;
4124 $sectionIndex =
false;
4126 $markerMatches = [];
4127 if ( preg_match(
"/^$markerRegex/", $headline, $markerMatches ) ) {
4128 $serial = $markerMatches[1];
4129 list( $titleText, $sectionIndex ) = $this->mHeadings[$serial];
4130 $isTemplate = ( $titleText != $baseTitleText );
4131 $headline = preg_replace(
"/^$markerRegex\\s*/",
"", $headline );
4135 $prevlevel = $level;
4137 $level =
$matches[1][$headlineCount];
4139 if ( $level > $prevlevel ) {
4140 # Increase TOC level
4142 $sublevelCount[$toclevel] = 0;
4143 if ( $toclevel < $wgMaxTocLevel ) {
4144 $prevtoclevel = $toclevel;
4148 } elseif ( $level < $prevlevel && $toclevel > 1 ) {
4149 # Decrease TOC level, find level to jump to
4151 for ( $i = $toclevel; $i > 0; $i-- ) {
4152 if ( $levelCount[$i] == $level ) {
4153 # Found last matching level
4156 } elseif ( $levelCount[$i] < $level ) {
4157 # Found first matching level below current level
4165 if ( $toclevel < $wgMaxTocLevel ) {
4166 if ( $prevtoclevel < $wgMaxTocLevel ) {
4167 # Unindent only if the previous toc level was shown :p
4169 $prevtoclevel = $toclevel;
4175 # No change in level, end TOC line
4176 if ( $toclevel < $wgMaxTocLevel ) {
4181 $levelCount[$toclevel] = $level;
4183 # count number of headlines for each level
4184 $sublevelCount[$toclevel]++;
4186 for ( $i = 1; $i <= $toclevel; $i++ ) {
4187 if ( !empty( $sublevelCount[$i] ) ) {
4191 $numbering .= $this->getTargetLanguage()->formatNum( $sublevelCount[$i] );
4196 # The safe header is a version of the header text safe to use for links
4198 # Remove link placeholders by the link text.
4199 # <!--LINK number-->
4201 # link text with suffix
4202 # Do this before unstrip since link text can contain strip markers
4203 $safeHeadline = $this->replaceLinkHoldersText( $headline );
4205 # Avoid insertion of weird stuff like <math> by expanding the relevant sections
4206 $safeHeadline = $this->mStripState->unstripBoth( $safeHeadline );
4208 # Strip out HTML (first regex removes any tag not allowed)
4210 # * <sup> and <sub> (T10393)
4214 # * <span dir="rtl"> and <span dir="ltr"> (T37167)
4215 # * <s> and <strike> (T35715)
4216 # We strip any parameter from accepted tags (second regex), except dir="rtl|ltr" from <span>,
4217 # to allow setting directionality in toc items.
4218 $tocline = preg_replace(
4220 '#<(?!/?(span|sup|sub|bdi|i|b|s|strike)(?: [^>]*)?>).*?>#',
4221 '#<(/?(?:span(?: dir="(?:rtl|ltr)")?|sup|sub|bdi|i|b|s|strike))(?: .*?)?>#'
4227 # Strip '<span></span>', which is the result from the above if
4228 # <span id="foo"></span> is used to produce an additional anchor
4230 $tocline = str_replace(
'<span></span>',
'', $tocline );
4232 $tocline = trim( $tocline );
4234 # For the anchor, strip out HTML-y stuff period
4235 $safeHeadline = preg_replace(
'/<.*?>/',
'', $safeHeadline );
4236 $safeHeadline = Sanitizer::normalizeSectionNameWhitespace( $safeHeadline );
4238 # Save headline for section edit hint before it's escaped
4239 $headlineHint = $safeHeadline;
4241 if ( $wgExperimentalHtmlIds ) {
4242 # For reverse compatibility, provide an id that's
4243 # HTML4-compatible, like we used to.
4244 # It may be worth noting, academically, that it's possible for
4245 # the legacy anchor to conflict with a non-legacy headline
4246 # anchor on the page. In this case likely the "correct" thing
4247 # would be to either drop the legacy anchors or make sure
4248 # they're numbered first. However, this would require people
4249 # to type in section names like "abc_.D7.93.D7.90.D7.A4"
4250 # manually, so let's not bother worrying about it.
4251 $legacyHeadline = Sanitizer::escapeId( $safeHeadline,
4252 [
'noninitial',
'legacy' ] );
4253 $safeHeadline = Sanitizer::escapeId( $safeHeadline );
4255 if ( $legacyHeadline == $safeHeadline ) {
4256 # No reason to have both (in fact, we can't)
4257 $legacyHeadline =
false;
4260 $legacyHeadline =
false;
4261 $safeHeadline = Sanitizer::escapeId( $safeHeadline,
4265 # HTML names must be case-insensitively unique (T12721).
4266 # This does not apply to Unicode characters per
4267 # https://www.w3.org/TR/html5/infrastructure.html#case-sensitivity-and-string-comparison
4268 # @todo FIXME: We may be changing them depending on the current locale.
4269 $arrayKey = strtolower( $safeHeadline );
4270 if ( $legacyHeadline ===
false ) {
4271 $legacyArrayKey =
false;
4273 $legacyArrayKey = strtolower( $legacyHeadline );
4276 # Create the anchor for linking from the TOC to the section
4277 $anchor = $safeHeadline;
4278 $legacyAnchor = $legacyHeadline;
4279 if ( isset( $refers[$arrayKey] ) ) {
4281 for ( $i = 2; isset( $refers[
"${arrayKey}_$i"] ); ++$i );
4284 $refers[
"${arrayKey}_$i"] =
true;
4286 $refers[$arrayKey] =
true;
4288 if ( $legacyHeadline !==
false && isset( $refers[$legacyArrayKey] ) ) {
4290 for ( $i = 2; isset( $refers[
"${legacyArrayKey}_$i"] ); ++$i );
4292 $legacyAnchor .=
"_$i";
4293 $refers[
"${legacyArrayKey}_$i"] =
true;
4295 $refers[$legacyArrayKey] =
true;
4298 # Don't number the heading if it is the only one (looks silly)
4299 if (
count(
$matches[3] ) > 1 && $this->mOptions->getNumberHeadings() ) {
4300 # the two are different if the line contains a link
4303 [
'class' =>
'mw-headline-number' ],
4305 ) .
' ' . $headline;
4308 if ( $enoughToc && ( !isset( $wgMaxTocLevel ) || $toclevel < $wgMaxTocLevel ) ) {
4310 $numbering, $toclevel, ( $isTemplate ?
false : $sectionIndex ) );
4313 # Add the section to the section tree
4314 # Find the DOM node for this header
4315 $noOffset = ( $isTemplate || $sectionIndex ===
false );
4316 while ( $node && !$noOffset ) {
4317 if ( $node->getName() ===
'h' ) {
4318 $bits = $node->splitHeading();
4319 if ( $bits[
'i'] == $sectionIndex ) {
4323 $byteOffset += mb_strlen( $this->mStripState->unstripBoth(
4325 $node = $node->getNextSibling();
4328 'toclevel' => $toclevel,
4331 'number' => $numbering,
4332 'index' => ( $isTemplate ?
'T-' :
'' ) . $sectionIndex,
4333 'fromtitle' => $titleText,
4334 'byteoffset' => ( $noOffset ?
null : $byteOffset ),
4335 'anchor' => $anchor,
4338 # give headline the correct <h#> tag
4339 if ( $maybeShowEditLink && $sectionIndex !==
false ) {
4341 if ( $isTemplate ) {
4342 # Put a T flag in the section identifier, to indicate to extractSections()
4343 # that sections inside <includeonly> should be counted.
4344 $editsectionPage = $titleText;
4345 $editsectionSection =
"T-$sectionIndex";
4346 $editsectionContent =
null;
4348 $editsectionPage = $this->mTitle->getPrefixedText();
4349 $editsectionSection = $sectionIndex;
4350 $editsectionContent = $headlineHint;
4364 $editlink =
'<mw:editsection page="' . htmlspecialchars( $editsectionPage );
4365 $editlink .=
'" section="' . htmlspecialchars( $editsectionSection ) .
'"';
4366 if ( $editsectionContent !==
null ) {
4367 $editlink .=
'>' . $editsectionContent .
'</mw:editsection>';
4375 $matches[
'attrib'][$headlineCount], $anchor, $headline,
4376 $editlink, $legacyAnchor );
4381 $this->setOutputType( $oldType );
4383 # Never ever show TOC if no headers
4384 if ( $numVisible < 1 ) {
4389 if ( $prevtoclevel > 0 && $prevtoclevel < $wgMaxTocLevel ) {
4393 $this->mOutput->setTOCHTML( $toc );
4394 $toc = self::TOC_START . $toc . self::TOC_END;
4395 $this->mOutput->addModules(
'mediawiki.toc' );
4399 $this->mOutput->setSections( $tocraw );
4402 # split up and insert constructed headlines
4403 $blocks = preg_split(
'/<H[1-6].*?>[\s\S]*?<\/H[1-6]>/i', $text );
4408 foreach ( $blocks
as $block ) {
4410 if ( empty( $head[$i - 1] ) ) {
4411 $sections[$i] = $block;
4413 $sections[$i] = $head[$i - 1] . $block;
4426 Hooks::run(
'ParserSectionCreate', [ $this, $i, &$sections[$i], $showEditLink ] );
4431 if ( $enoughToc && $isMain && !$this->mForceTocPosition ) {
4434 $sections[0] = $sections[0] . $toc .
"\n";
4437 $full .= implode(
'', $sections );
4439 if ( $this->mForceTocPosition ) {
4440 return str_replace(
'<!--MWTOC-->', $toc, $full );
4460 if ( $clearState ) {
4461 $magicScopeVariable = $this->lock();
4464 $this->setUser(
$user );
4467 $text = str_replace(
"\000",
'', $text );
4474 if (
$options->getPreSaveTransform() ) {
4475 $text = $this->pstPass2( $text,
$user );
4477 $text = $this->mStripState->unstripBoth( $text );
4479 $this->setUser(
null ); # Reset
4492 private function pstPass2( $text,
$user ) {
4495 # Note: This is the timestamp saved as hardcoded wikitext to
4496 # the database, we use $wgContLang here in order to give
4497 # everyone the same signature and use the default one rather
4498 # than the one selected in each user's preferences.
4500 $ts = $this->mOptions->getTimestamp();
4502 $ts = $timestamp->format(
'YmdHis' );
4503 $tzMsg = $timestamp->getTimezoneMessage()->inContentLanguage()->text();
4505 $d =
$wgContLang->timeanddate( $ts,
false,
false ) .
" ($tzMsg)";
4507 # Variable replacement
4508 # Because mOutputType is OT_WIKI, this will only process {{subst:xxx}} type tags
4509 $text = $this->replaceVariables( $text );
4511 # This works almost by chance, as the replaceVariables are done before the getUserSig(),
4512 # which may corrupt this parser instance via its wfMessage()->text() call-
4515 $sigText = $this->getUserSig(
$user );
4516 $text = strtr( $text, [
4518 '~~~~' =>
"$sigText $d",
4522 # Context links ("pipe tricks"): [[|name]] and [[name (context)|]]
4524 $nc =
'[ _0-9A-Za-z\x80-\xff-]'; # Namespaces can
use non-ascii!
4527 $p1 =
"/\[\[(:?$nc+:|:|)($tc+?)( ?\\($tc+\\))\\|]]/";
4529 $p4 =
"/\[\[(:?$nc+:|:|)($tc+?)( ?($tc+))\\|]]/";
4531 $p3 =
"/\[\[(:?$nc+:|:|)($tc+?)( ?\\($tc+\\)|)((?:, |,)$tc+|)\\|]]/";
4533 $p2 =
"/\[\[\\|($tc+)]]/";
4535 # try $p1 first, to turn "[[A, B (C)|]]" into "[[A, B (C)|A, B]]"
4536 $text = preg_replace( $p1,
'[[\\1\\2\\3|\\2]]', $text );
4537 $text = preg_replace( $p4,
'[[\\1\\2\\3|\\2]]', $text );
4538 $text = preg_replace( $p3,
'[[\\1\\2\\3\\4|\\2]]', $text );
4540 $t = $this->mTitle->getText();
4542 if ( preg_match(
"/^($nc+:|)$tc+?( \\($tc+\\))$/",
$t, $m ) ) {
4543 $text = preg_replace( $p2,
"[[$m[1]\\1$m[2]|\\1]]", $text );
4544 } elseif ( preg_match(
"/^($nc+:|)$tc+?(, $tc+|)$/",
$t, $m ) &&
"$m[1]$m[2]" !=
'' ) {
4545 $text = preg_replace( $p2,
"[[$m[1]\\1$m[2]|\\1]]", $text );
4547 # if there's no context, don't bother duplicating the title
4548 $text = preg_replace( $p2,
'[[\\1]]', $text );
4568 public function getUserSig( &
$user, $nickname =
false, $fancySig =
null ) {
4573 # If not given, retrieve from the user object.
4574 if ( $nickname ===
false ) {
4575 $nickname =
$user->getOption(
'nickname' );
4578 if ( is_null( $fancySig ) ) {
4579 $fancySig =
$user->getBoolOption(
'fancysig' );
4582 $nickname = $nickname ==
null ?
$username : $nickname;
4584 if ( mb_strlen( $nickname ) > $wgMaxSigChars ) {
4586 wfDebug( __METHOD__ .
": $username has overlong signature.\n" );
4587 } elseif ( $fancySig !==
false ) {
4588 # Sig. might contain markup; validate this
4589 if ( $this->validateSig( $nickname ) !==
false ) {
4590 # Validated; clean up (if needed) and return it
4591 return $this->cleanSig( $nickname,
true );
4593 # Failed to validate; fall back to the default
4595 wfDebug( __METHOD__ .
": $username has bad XML tags in signature.\n" );
4599 # Make sure nickname doesnt get a sig in a sig
4600 $nickname = self::cleanSigInSig( $nickname );
4602 # If we're still here, make it a link to the user page
4605 $msgName =
$user->isAnon() ?
'signature-anon' :
'signature';
4607 return wfMessage( $msgName, $userText, $nickText )->inContentLanguage()
4608 ->title( $this->getTitle() )->text();
4617 public function validateSig( $text ) {
4631 public function cleanSig( $text, $parsing =
false ) {
4634 $magicScopeVariable = $this->lock();
4638 # Option to disable this feature
4639 if ( !$this->mOptions->getCleanSignatures() ) {
4643 # @todo FIXME: Regex doesn't respect extension tags or nowiki
4644 # => Move this logic to braceSubstitution()
4646 $substRegex =
'/\{\{(?!(?:' . $substWord->getBaseRegex() .
'))/x' . $substWord->getRegexCase();
4647 $substText =
'{{' . $substWord->getSynonym( 0 );
4649 $text = preg_replace( $substRegex, $substText, $text );
4650 $text = self::cleanSigInSig( $text );
4651 $dom = $this->preprocessToDom( $text );
4652 $frame = $this->getPreprocessor()->newFrame();
4653 $text = $frame->expand( $dom );
4656 $text = $this->mStripState->unstripBoth( $text );
4668 public static function cleanSigInSig( $text ) {
4669 $text = preg_replace(
'/~{3,5}/',
'', $text );
4683 $outputType, $clearState =
true
4695 $outputType, $clearState =
true
4697 $this->setTitle(
$title );
4699 $this->setOutputType( $outputType );
4700 if ( $clearState ) {
4701 $this->clearState();
4714 static $executing =
false;
4716 # Guard against infinite recursion
4757 public function setHook(
$tag, $callback ) {
4759 if ( preg_match(
'/[<>\r\n]/',
$tag, $m ) ) {
4760 throw new MWException(
"Invalid character {$m[0]} in setHook('$tag', ...) call" );
4762 $oldVal = isset( $this->mTagHooks[
$tag] ) ? $this->mTagHooks[
$tag] :
null;
4763 $this->mTagHooks[
$tag] = $callback;
4764 if ( !in_array(
$tag, $this->mStripList ) ) {
4765 $this->mStripList[] =
$tag;
4788 public function setTransparentTagHook(
$tag, $callback ) {
4790 if ( preg_match(
'/[<>\r\n]/',
$tag, $m ) ) {
4791 throw new MWException(
"Invalid character {$m[0]} in setTransparentHook('$tag', ...) call" );
4793 $oldVal = isset( $this->mTransparentTagHooks[
$tag] ) ? $this->mTransparentTagHooks[
$tag] :
null;
4794 $this->mTransparentTagHooks[
$tag] = $callback;
4802 public function clearTagHooks() {
4803 $this->mTagHooks = [];
4804 $this->mFunctionTagHooks = [];
4805 $this->mStripList = $this->mDefaultStripList;
4851 public function setFunctionHook( $id, $callback,
$flags = 0 ) {
4854 $oldVal = isset( $this->mFunctionHooks[$id] ) ? $this->mFunctionHooks[$id][0] :
null;
4855 $this->mFunctionHooks[$id] = [ $callback,
$flags ];
4857 # Add to function cache
4860 throw new MWException( __METHOD__ .
'() expecting a magic word identifier.' );
4863 $synonyms = $mw->getSynonyms();
4864 $sensitive = intval( $mw->isCaseSensitive() );
4866 foreach ( $synonyms
as $syn ) {
4868 if ( !$sensitive ) {
4875 # Remove trailing colon
4876 if ( substr( $syn, -1, 1 ) ===
':' ) {
4877 $syn = substr( $syn, 0, -1 );
4879 $this->mFunctionSynonyms[$sensitive][$syn] = $id;
4889 public function getFunctionHooks() {
4890 return array_keys( $this->mFunctionHooks );
4903 public function setFunctionTagHook(
$tag, $callback,
$flags ) {
4905 if ( preg_match(
'/[<>\r\n]/',
$tag, $m ) ) {
4906 throw new MWException(
"Invalid character {$m[0]} in setFunctionTagHook('$tag', ...) call" );
4908 $old = isset( $this->mFunctionTagHooks[
$tag] ) ?
4909 $this->mFunctionTagHooks[
$tag] :
null;
4910 $this->mFunctionTagHooks[
$tag] = [ $callback,
$flags ];
4912 if ( !in_array(
$tag, $this->mStripList ) ) {
4913 $this->mStripList[] =
$tag;
4926 public function replaceLinkHolders( &$text,
$options = 0 ) {
4927 $this->mLinkHolders->replace( $text );
4937 public function replaceLinkHoldersText( $text ) {
4938 return $this->mLinkHolders->replaceText( $text );
4954 public function renderImageGallery( $text,
$params ) {
4957 if ( isset(
$params[
'mode'] ) ) {
4963 }
catch ( Exception
$e ) {
4968 $ig->setContextTitle( $this->mTitle );
4969 $ig->setShowBytes(
false );
4970 $ig->setShowFilename(
false );
4971 $ig->setParser( $this );
4972 $ig->setHideBadImages();
4973 $ig->setAttributes( Sanitizer::validateTagAttributes(
$params,
'ul' ) );
4975 if ( isset(
$params[
'showfilename'] ) ) {
4976 $ig->setShowFilename(
true );
4978 $ig->setShowFilename(
false );
4980 if ( isset(
$params[
'caption'] ) ) {
4981 $caption =
$params[
'caption'];
4982 $caption = htmlspecialchars( $caption );
4983 $caption = $this->replaceInternalLinks( $caption );
4984 $ig->setCaptionHtml( $caption );
4986 if ( isset(
$params[
'perrow'] ) ) {
4987 $ig->setPerRow(
$params[
'perrow'] );
4989 if ( isset(
$params[
'widths'] ) ) {
4990 $ig->setWidths(
$params[
'widths'] );
4992 if ( isset(
$params[
'heights'] ) ) {
4993 $ig->setHeights(
$params[
'heights'] );
4995 $ig->setAdditionalOptions(
$params );
5003 # match lines like these:
5004 # Image:someimage.jpg|This is some image
5012 if ( strpos(
$matches[0],
'%' ) !==
false ) {
5016 if ( is_null(
$title ) ) {
5017 # Bogus title. Ignore these so we don't bomb out later.
5021 # We need to get what handler the file uses, to figure out parameters.
5022 # Note, a hook can overide the file name, and chose an entirely different
5023 # file (which potentially could be of a different type and have different handler).
5028 # Don't register it now, as TraditionalImageGallery does that later.
5030 $handler = $file ? $file->getHandler() :
false;
5033 'img_alt' =>
'gallery-internal-alt',
5034 'img_link' =>
'gallery-internal-link',
5037 $paramMap = $paramMap +
$handler->getParamMap();
5040 unset( $paramMap[
'img_width'] );
5048 $handlerOptions = [];
5062 foreach ( $parameterMatches
as $parameterMatch ) {
5063 list( $magicName, $match ) = $mwArray->matchVariableStartToEnd( $parameterMatch );
5065 $paramName = $paramMap[$magicName];
5067 switch ( $paramName ) {
5068 case 'gallery-internal-alt':
5069 $alt = $this->stripAltText( $match,
false );
5071 case 'gallery-internal-link':
5072 $linkValue = strip_tags( $this->replaceLinkHoldersText( $match ) );
5073 $chars = self::EXT_LINK_URL_CLASS;
5074 $addr = self::EXT_LINK_ADDR;
5075 $prots = $this->mUrlProtocols;
5077 if ( preg_match(
'/^-{R|(.*)}-$/', $linkValue ) ) {
5080 $linkValue = substr( $linkValue, 4, -2 );
5082 if ( preg_match(
"/^($prots)$addr$chars*$/u", $linkValue ) ) {
5084 $this->mOutput->addExternalLink(
$link );
5087 if ( $localLinkTitle !==
null ) {
5088 $this->mOutput->addLink( $localLinkTitle );
5089 $link = $localLinkTitle->getLinkURL();
5095 if (
$handler->validateParam( $paramName, $match ) ) {
5096 $handlerOptions[$paramName] = $match;
5099 wfDebug(
"$parameterMatch failed parameter validation\n" );
5100 $label =
'|' . $parameterMatch;
5106 $label =
'|' . $parameterMatch;
5110 $label = substr( $label, 1 );
5113 $ig->add(
$title, $label, $alt,
$link, $handlerOptions );
5115 $html = $ig->toHTML();
5124 public function getImageParams(
$handler ) {
5126 $handlerClass = get_class(
$handler );
5130 if ( !isset( $this->mImageParams[$handlerClass] ) ) {
5131 # Initialise static lists
5132 static $internalParamNames = [
5133 'horizAlign' => [
'left',
'right',
'center',
'none' ],
5134 'vertAlign' => [
'baseline',
'sub',
'super',
'top',
'text-top',
'middle',
5135 'bottom',
'text-bottom' ],
5136 'frame' => [
'thumbnail',
'manualthumb',
'framed',
'frameless',
5137 'upright',
'border',
'link',
'alt',
'class' ],
5139 static $internalParamMap;
5140 if ( !$internalParamMap ) {
5141 $internalParamMap = [];
5142 foreach ( $internalParamNames
as $type => $names ) {
5149 $magicName = str_replace(
'-',
'_',
"img_$name" );
5150 $internalParamMap[$magicName] = [
$type,
$name ];
5155 # Add handler params
5156 $paramMap = $internalParamMap;
5158 $handlerParamMap =
$handler->getParamMap();
5159 foreach ( $handlerParamMap
as $magic => $paramName ) {
5160 $paramMap[$magic] = [
'handler', $paramName ];
5163 $this->mImageParams[$handlerClass] = $paramMap;
5164 $this->mImageParamsMagicArray[$handlerClass] =
new MagicWordArray( array_keys( $paramMap ) );
5166 return [ $this->mImageParams[$handlerClass], $this->mImageParamsMagicArray[$handlerClass] ];
5177 public function makeImage(
$title,
$options, $holders =
false ) {
5178 # Check if the options text is of the form "options|alt text"
5180 # * thumbnail make a thumbnail with enlarge-icon and caption, alignment depends on lang
5181 # * left no resizing, just left align. label is used for alt= only
5182 # * right same, but right aligned
5183 # * none same, but not aligned
5184 # * ___px scale to ___ pixels width, no aligning. e.g. use in taxobox
5185 # * center center the image
5186 # * frame Keep original image size, no magnify-button.
5187 # * framed Same as "frame"
5188 # * frameless like 'thumb' but without a frame. Keeps user preferences for width
5189 # * upright reduce width for upright images, rounded to full __0 px
5190 # * border draw a 1px border around the image
5191 # * alt Text for HTML alt attribute (defaults to empty)
5192 # * class Set a class for img node
5193 # * link Set the target of the image link. Can be external, interwiki, or local
5194 # vertical-align values (no % or length right now):
5204 # Protect LanguageConverter markup when splitting into parts
5209 # Give extensions a chance to select the file revision for us
5214 # Fetch and register the file (file title may be different via hooks)
5218 $handler = $file ? $file->getHandler() :
false;
5220 list( $paramMap, $mwArray ) = $this->getImageParams(
$handler );
5223 $this->addTrackingCategory(
'broken-file-category' );
5226 # Process the input parameters
5228 $params = [
'frame' => [],
'handler' => [],
5229 'horizAlign' => [],
'vertAlign' => [] ];
5230 $seenformat =
false;
5231 foreach ( $parts
as $part ) {
5232 $part = trim( $part );
5233 list( $magicName,
$value ) = $mwArray->matchVariableStartToEnd( $part );
5235 if ( isset( $paramMap[$magicName] ) ) {
5236 list(
$type, $paramName ) = $paramMap[$magicName];
5238 # Special case; width and height come in one variable together
5239 if (
$type ===
'handler' && $paramName ===
'width' ) {
5240 $parsedWidthParam = $this->parseWidthParam(
$value );
5241 if ( isset( $parsedWidthParam[
'width'] ) ) {
5242 $width = $parsedWidthParam[
'width'];
5243 if (
$handler->validateParam(
'width', $width ) ) {
5248 if ( isset( $parsedWidthParam[
'height'] ) ) {
5249 $height = $parsedWidthParam[
'height'];
5250 if (
$handler->validateParam(
'height', $height ) ) {
5255 # else no validation -- T15436
5257 if (
$type ===
'handler' ) {
5258 # Validate handler parameter
5261 # Validate internal parameters
5262 switch ( $paramName ) {
5266 # @todo FIXME: Possibly check validity here for
5267 # manualthumb? downstream behavior seems odd with
5268 # missing manual thumbs.
5273 $chars = self::EXT_LINK_URL_CLASS;
5274 $addr = self::EXT_LINK_ADDR;
5275 $prots = $this->mUrlProtocols;
5277 $paramName =
'no-link';
5280 } elseif ( preg_match(
"/^((?i)$prots)/",
$value ) ) {
5281 if ( preg_match(
"/^((?i)$prots)$addr$chars*$/u",
$value, $m ) ) {
5282 $paramName =
'link-url';
5283 $this->mOutput->addExternalLink(
$value );
5284 if ( $this->mOptions->getExternalLinkTarget() ) {
5285 $params[
$type][
'link-target'] = $this->mOptions->getExternalLinkTarget();
5292 $paramName =
'link-title';
5294 $this->mOutput->addLink( $linkTitle );
5303 $validated = !$seenformat;
5307 # Most other things appear to be empty or numeric...
5308 $validated = (
$value ===
false || is_numeric( trim(
$value ) ) );
5317 if ( !$validated ) {
5322 # Process alignment parameters
5323 if (
$params[
'horizAlign'] ) {
5330 $params[
'frame'][
'caption'] = $caption;
5332 # Will the image be presented in a frame, with the caption below?
5333 $imageIsFramed = isset(
$params[
'frame'][
'frame'] )
5334 || isset(
$params[
'frame'][
'framed'] )
5335 || isset(
$params[
'frame'][
'thumbnail'] )
5336 || isset(
$params[
'frame'][
'manualthumb'] );
5338 # In the old days, [[Image:Foo|text...]] would set alt text. Later it
5339 # came to also set the caption, ordinary text after the image -- which
5340 # makes no sense, because that just repeats the text multiple times in
5341 # screen readers. It *also* came to set the title attribute.
5342 # Now that we have an alt attribute, we should not set the alt text to
5343 # equal the caption: that's worse than useless, it just repeats the
5344 # text. This is the framed/thumbnail case. If there's no caption, we
5345 # use the unnamed parameter for alt text as well, just for the time be-
5346 # ing, if the unnamed param is set and the alt param is not.
5347 # For the future, we need to figure out if we want to tweak this more,
5348 # e.g., introducing a title= parameter for the title; ignoring the un-
5349 # named parameter entirely for images without a caption; adding an ex-
5350 # plicit caption= parameter and preserving the old magic unnamed para-
5352 if ( $imageIsFramed ) { # Framed image
5353 if ( $caption ===
'' && !isset(
$params[
'frame'][
'alt'] ) ) {
5354 # No caption or alt text, add the filename as the alt text so
5355 # that screen readers at least get some description of the image
5358 # Do not set $params['frame']['title'] because tooltips don't make sense
5360 }
else { # Inline image
5361 if ( !isset(
$params[
'frame'][
'alt'] ) ) {
5362 # No alt text, use the "caption" for the alt text
5363 if ( $caption !==
'' ) {
5364 $params[
'frame'][
'alt'] = $this->stripAltText( $caption, $holders );
5366 # No caption, fall back to using the filename for the
5371 # Use the "caption" for the tooltip text
5372 $params[
'frame'][
'title'] = $this->stripAltText( $caption, $holders );
5377 # Linker does the rest
5380 $time, $descQuery, $this->mOptions->getThumbSize() );
5382 # Give the handler a chance to modify the parser object
5384 $handler->parserTransformHook( $this, $file );
5395 protected function stripAltText( $caption, $holders ) {
5396 # Strip bad stuff out of the title (tooltip). We can't just use
5397 # replaceLinkHoldersText() here, because if this function is called
5398 # from replaceInternalLinks2(), mLinkHolders won't be up-to-date.
5400 $tooltip = $holders->replaceText( $caption );
5402 $tooltip = $this->replaceLinkHoldersText( $caption );
5405 # make sure there are no placeholders in thumbnail attributes
5406 # that are later expanded to html- so expand them now and
5408 $tooltip = $this->mStripState->unstripBoth( $tooltip );
5409 $tooltip = Sanitizer::stripAllTags( $tooltip );
5419 public function disableCache() {
5420 wfDebug(
"Parser output marked as uncacheable.\n" );
5421 if ( !$this->mOutput ) {
5423 " can only be called when actually parsing something" );
5425 $this->mOutput->updateCacheExpiry( 0 );
5436 public function attributeStripCallback( &$text, $frame =
false ) {
5437 $text = $this->replaceVariables( $text, $frame );
5438 $text = $this->mStripState->unstripBoth( $text );
5447 public function getTags() {
5449 array_keys( $this->mTransparentTagHooks ),
5450 array_keys( $this->mTagHooks ),
5451 array_keys( $this->mFunctionTagHooks )
5465 public function replaceTransparentTags( $text ) {
5467 $elements = array_keys( $this->mTransparentTagHooks );
5468 $text = self::extractTagsAndParams( $elements, $text,
$matches );
5473 $tagName = strtolower( $element );
5474 if ( isset( $this->mTransparentTagHooks[$tagName] ) ) {
5475 $output = call_user_func_array(
5476 $this->mTransparentTagHooks[$tagName],
5482 $replacements[$marker] =
$output;
5484 return strtr( $text, $replacements );
5516 private function extractSections( $text, $sectionId, $mode, $newText =
'' ) {
5519 $magicScopeVariable = $this->lock();
5522 $frame = $this->getPreprocessor()->newFrame();
5524 # Process section extraction flags
5526 $sectionParts = explode(
'-', $sectionId );
5527 $sectionIndex = array_pop( $sectionParts );
5528 foreach ( $sectionParts
as $part ) {
5529 if ( $part ===
'T' ) {
5530 $flags |= self::PTD_FOR_INCLUSION;
5534 # Check for empty input
5535 if ( strval( $text ) ===
'' ) {
5536 # Only sections 0 and T-0 exist in an empty document
5537 if ( $sectionIndex == 0 ) {
5538 if ( $mode ===
'get' ) {
5544 if ( $mode ===
'get' ) {
5552 # Preprocess the text
5553 $root = $this->preprocessToDom( $text,
$flags );
5555 # <h> nodes indicate section breaks
5556 # They can only occur at the top level, so we can find them by iterating the root's children
5557 $node = $root->getFirstChild();
5559 # Find the target section
5560 if ( $sectionIndex == 0 ) {
5561 # Section zero doesn't nest, level=big
5562 $targetLevel = 1000;
5565 if ( $node->getName() ===
'h' ) {
5566 $bits = $node->splitHeading();
5567 if ( $bits[
'i'] == $sectionIndex ) {
5568 $targetLevel = $bits[
'level'];
5572 if ( $mode ===
'replace' ) {
5575 $node = $node->getNextSibling();
5581 if ( $mode ===
'get' ) {
5588 # Find the end of the section, including nested sections
5590 if ( $node->getName() ===
'h' ) {
5591 $bits = $node->splitHeading();
5592 $curLevel = $bits[
'level'];
5593 if ( $bits[
'i'] != $sectionIndex && $curLevel <= $targetLevel ) {
5597 if ( $mode ===
'get' ) {
5600 $node = $node->getNextSibling();
5603 # Write out the remainder (in replace mode only)
5604 if ( $mode ===
'replace' ) {
5605 # Output the replacement text
5606 # Add two newlines on -- trailing whitespace in $newText is conventionally
5607 # stripped by the editor, so we need both newlines to restore the paragraph gap
5608 # Only add trailing whitespace if there is newText
5609 if ( $newText !=
"" ) {
5610 $outText .= $newText .
"\n\n";
5615 $node = $node->getNextSibling();
5619 if ( is_string( $outText ) ) {
5620 # Re-insert stripped tags
5621 $outText = rtrim( $this->mStripState->unstripBoth( $outText ) );
5641 public function getSection( $text, $sectionId, $defaultText =
'' ) {
5642 return $this->extractSections( $text, $sectionId,
'get', $defaultText );
5657 public function replaceSection( $oldText, $sectionId, $newText ) {
5658 return $this->extractSections( $oldText, $sectionId,
'replace', $newText );
5666 public function getRevisionId() {
5667 return $this->mRevisionId;
5676 public function getRevisionObject() {
5677 if ( !is_null( $this->mRevisionObject ) ) {
5678 return $this->mRevisionObject;
5680 if ( is_null( $this->mRevisionId ) ) {
5684 $rev = call_user_func(
5685 $this->mOptions->getCurrentRevisionCallback(), $this->getTitle(), $this
5688 # If the parse is for a new revision, then the callback should have
5689 # already been set to force the object and should match mRevisionId.
5690 # If not, try to fetch by mRevisionId for sanity.
5691 if (
$rev &&
$rev->getId() != $this->mRevisionId ) {
5695 $this->mRevisionObject =
$rev;
5697 return $this->mRevisionObject;
5705 public function getRevisionTimestamp() {
5706 if ( is_null( $this->mRevisionTimestamp ) ) {
5709 $revObject = $this->getRevisionObject();
5710 $timestamp = $revObject ? $revObject->getTimestamp() :
wfTimestampNow();
5712 # The cryptic '' timezone parameter tells to use the site-default
5713 # timezone offset instead of the user settings.
5714 # Since this value will be saved into the parser cache, served
5715 # to other users, and potentially even used inside links and such,
5716 # it needs to be consistent for all visitors.
5717 $this->mRevisionTimestamp =
$wgContLang->userAdjust( $timestamp,
'' );
5720 return $this->mRevisionTimestamp;
5728 public function getRevisionUser() {
5729 if ( is_null( $this->mRevisionUser ) ) {
5730 $revObject = $this->getRevisionObject();
5732 # if this template is subst: the revision id will be blank,
5733 # so just use the current user's name
5735 $this->mRevisionUser = $revObject->getUserText();
5736 } elseif ( $this->ot[
'wiki'] || $this->mOptions->getIsPreview() ) {
5737 $this->mRevisionUser = $this->getUser()->getName();
5740 return $this->mRevisionUser;
5748 public function getRevisionSize() {
5749 if ( is_null( $this->mRevisionSize ) ) {
5750 $revObject = $this->getRevisionObject();
5752 # if this variable is subst: the revision id will be blank,
5753 # so just use the parser input size, because the own substituation
5754 # will change the size.
5756 $this->mRevisionSize = $revObject->getSize();
5758 $this->mRevisionSize = $this->mInputSize;
5761 return $this->mRevisionSize;
5769 public function setDefaultSort(
$sort ) {
5770 $this->mDefaultSort =
$sort;
5771 $this->mOutput->setProperty(
'defaultsort',
$sort );
5784 public function getDefaultSort() {
5785 if ( $this->mDefaultSort !==
false ) {
5786 return $this->mDefaultSort;
5798 public function getCustomDefaultSort() {
5799 return $this->mDefaultSort;
5811 public function guessSectionNameFromWikiText( $text ) {
5812 # Strip out wikitext links(they break the anchor)
5813 $text = $this->stripSectionName( $text );
5814 $text = Sanitizer::normalizeSectionNameWhitespace( $text );
5815 return '#' . Sanitizer::escapeId( $text,
'noninitial' );
5826 public function guessLegacySectionNameFromWikiText( $text ) {
5827 # Strip out wikitext links(they break the anchor)
5828 $text = $this->stripSectionName( $text );
5829 $text = Sanitizer::normalizeSectionNameWhitespace( $text );
5830 return '#' . Sanitizer::escapeId( $text, [
'noninitial',
'legacy' ] );
5847 public function stripSectionName( $text ) {
5848 # Strip internal link markup
5849 $text = preg_replace(
'/\[\[:?([^[|]+)\|([^[]+)\]\]/',
'$2', $text );
5850 $text = preg_replace(
'/\[\[:?([^[]+)\|?\]\]/',
'$1', $text );
5852 # Strip external link markup
5853 # @todo FIXME: Not tolerant to blank link text
5854 # I.E. [https://www.mediawiki.org] will render as [1] or something depending
5855 # on how many empty links there are on the page - need to figure that out.
5856 $text = preg_replace(
'/\[(?i:' . $this->mUrlProtocols .
')([^ ]+?) ([^[]+)\]/',
'$2', $text );
5858 # Parse wikitext quotes (italics & bold)
5859 $text = $this->doQuotes( $text );
5879 $magicScopeVariable = $this->lock();
5882 $text = $this->replaceVariables( $text );
5883 $text = $this->mStripState->unstripBoth( $text );
5884 $text = Sanitizer::removeHTMLtags( $text );
5924 public function markerSkipCallback(
$s, $callback ) {
5927 while ( $i < strlen(
$s ) ) {
5928 $markerStart = strpos(
$s, self::MARKER_PREFIX, $i );
5929 if ( $markerStart ===
false ) {
5930 $out .= call_user_func( $callback, substr(
$s, $i ) );
5933 $out .= call_user_func( $callback, substr(
$s, $i, $markerStart - $i ) );
5934 $markerEnd = strpos(
$s, self::MARKER_SUFFIX, $markerStart );
5935 if ( $markerEnd ===
false ) {
5936 $out .= substr(
$s, $markerStart );
5939 $markerEnd += strlen( self::MARKER_SUFFIX );
5940 $out .= substr(
$s, $markerStart, $markerEnd - $markerStart );
5954 public function killMarkers( $text ) {
5955 return $this->mStripState->killMarkers( $text );
5974 public function serializeHalfParsedText( $text ) {
5977 'version' => self::HALF_PARSED_VERSION,
5978 'stripState' => $this->mStripState->getSubState( $text ),
5979 'linkHolders' => $this->mLinkHolders->getSubArray( $text )
5999 public function unserializeHalfParsedText( $data ) {
6000 if ( !isset( $data[
'version'] ) || $data[
'version'] != self::HALF_PARSED_VERSION ) {
6001 throw new MWException( __METHOD__ .
': invalid version' );
6004 # First, extract the strip state.
6005 $texts = [ $data[
'text'] ];
6006 $texts = $this->mStripState->merge( $data[
'stripState'], $texts );
6008 # Now renumber links
6009 $texts = $this->mLinkHolders->mergeForeign( $data[
'linkHolders'], $texts );
6011 # Should be good to go.
6024 public function isValidHalfParsedText( $data ) {
6025 return isset( $data[
'version'] ) && $data[
'version'] == self::HALF_PARSED_VERSION;
6036 public function parseWidthParam(
$value ) {
6037 $parsedWidthParam = [];
6039 return $parsedWidthParam;
6042 # (T15500) In both cases (width/height and width only),
6043 # permit trailing "px" for backward compatibility.
6044 if ( preg_match(
'/^([0-9]*)x([0-9]*)\s*(?:px)?\s*$/',
$value, $m ) ) {
6045 $width = intval( $m[1] );
6046 $height = intval( $m[2] );
6047 $parsedWidthParam[
'width'] = $width;
6048 $parsedWidthParam[
'height'] = $height;
6049 } elseif ( preg_match(
'/^[0-9]*\s*(?:px)?\s*$/',
$value ) ) {
6050 $width = intval(
$value );
6051 $parsedWidthParam[
'width'] = $width;
6053 return $parsedWidthParam;
6065 protected function lock() {
6066 if ( $this->mInParse ) {
6067 throw new MWException(
"Parser state cleared while parsing. "
6068 .
"Did you call Parser::parse recursively?" );
6070 $this->mInParse =
true;
6072 $recursiveCheck =
new ScopedCallback(
function() {
6073 $this->mInParse =
false;
6076 return $recursiveCheck;
6089 public static function stripOuterParagraph(
$html ) {
6091 if ( preg_match(
'/^<p>(.*)\n?<\/p>\n?$/sU',
$html, $m ) ) {
6092 if ( strpos( $m[1],
'</p>' ) ===
false ) {
6110 public function getFreshParser() {
6112 if ( $this->mInParse ) {
6113 return new $wgParserConf[
'class']( $wgParserConf );
6125 public function enableOOUI() {
6127 $this->mOutput->setEnableOOUI(
true );