}
/**
- * Cleans up HTML, removes dangerous tags and attributes, and
- * removes HTML comments
- * @param string $text
- * @param callable $processCallback Callback to do any variable or parameter
- * replacements in HTML attribute values
- * @param array|bool $args Arguments for the processing callback
+ * Return the various lists of recognized tags
* @param array $extratags For any extra tags to include
* @param array $removetags For any tags (default or extra) to exclude
- * @return string
+ * @return array
*/
- public static function removeHTMLtags( $text, $processCallback = null,
- $args = array(), $extratags = array(), $removetags = array()
- ) {
- global $wgUseTidy, $wgAllowMicrodataAttributes, $wgAllowImageTag;
+ public static function getRecognizedTagData( $extratags = array(), $removetags = array() ) {
+ global $wgAllowMicrodataAttributes, $wgAllowImageTag;
static $htmlpairsStatic, $htmlsingle, $htmlsingleonly, $htmlnest, $tabletags,
$htmllist, $listtags, $htmlsingleallowed, $htmlelementsStatic, $staticInitialised;
// are changed (like in the screwed up test system) we will re-initialise the settings.
$globalContext = implode( '-', compact( 'wgAllowMicrodataAttributes', 'wgAllowImageTag' ) );
if ( !$staticInitialised || $staticInitialised != $globalContext ) {
-
$htmlpairsStatic = array( # Tags that must be closed
'b', 'bdi', 'del', 'i', 'ins', 'u', 'font', 'big', 'small', 'sub', 'sup', 'h1',
'h2', 'h3', 'h4', 'h5', 'h6', 'cite', 'code', 'em', 's',
}
$staticInitialised = $globalContext;
}
+
# Populate $htmlpairs and $htmlelements with the $extratags and $removetags arrays
$extratags = array_flip( $extratags );
$removetags = array_flip( $removetags );
$htmlpairs = array_merge( $extratags, $htmlpairsStatic );
$htmlelements = array_diff_key( array_merge( $extratags, $htmlelementsStatic ), $removetags );
+ return array(
+ 'htmlpairs' => $htmlpairs,
+ 'htmlsingle' => $htmlsingle,
+ 'htmlsingleonly' => $htmlsingleonly,
+ 'htmlnest' => $htmlnest,
+ 'tabletags' => $tabletags,
+ 'htmllist' => $htmllist,
+ 'listtags' => $listtags,
+ 'htmlsingleallowed' => $htmlsingleallowed,
+ 'htmlelements' => $htmlelements,
+ );
+ }
+
+ /**
+ * Cleans up HTML, removes dangerous tags and attributes, and
+ * removes HTML comments
+ * @param string $text
+ * @param callable $processCallback Callback to do any variable or parameter
+ * replacements in HTML attribute values
+ * @param array|bool $args Arguments for the processing callback
+ * @param array $extratags For any extra tags to include
+ * @param array $removetags For any tags (default or extra) to exclude
+ * @return string
+ */
+ public static function removeHTMLtags( $text, $processCallback = null,
+ $args = array(), $extratags = array(), $removetags = array()
+ ) {
+ global $wgUseTidy;
+
+ extract( self::getRecognizedTagData( $extratags, $removetags ) );
+
# Remove HTML comments
$text = Sanitizer::removeHTMLcomments( $text );
$bits = explode( '<', $text );
$badtag = true;
} elseif ( $slash ) {
# Closing a tag... is it the one we just opened?
- wfSuppressWarnings();
+ MediaWiki\suppressWarnings();
$ot = array_pop( $tagstack );
- wfRestoreWarnings();
+ MediaWiki\restoreWarnings();
if ( $ot != $t ) {
if ( isset( $htmlsingleallowed[$ot] ) ) {
# and see if we find a match below them
$optstack = array();
array_push( $optstack, $ot );
- wfSuppressWarnings();
+ MediaWiki\suppressWarnings();
$ot = array_pop( $tagstack );
- wfRestoreWarnings();
+ MediaWiki\restoreWarnings();
while ( $ot != $t && isset( $htmlsingleallowed[$ot] ) ) {
array_push( $optstack, $ot );
- wfSuppressWarnings();
+ MediaWiki\suppressWarnings();
$ot = array_pop( $tagstack );
- wfRestoreWarnings();
+ MediaWiki\restoreWarnings();
}
if ( $t != $ot ) {
# No match. Push the optional elements back again
$badtag = true;
- wfSuppressWarnings();
+ MediaWiki\suppressWarnings();
$ot = array_pop( $optstack );
- wfRestoreWarnings();
+ MediaWiki\restoreWarnings();
while ( $ot ) {
array_push( $tagstack, $ot );
- wfSuppressWarnings();
+ MediaWiki\suppressWarnings();
$ot = array_pop( $optstack );
- wfRestoreWarnings();
+ MediaWiki\restoreWarnings();
}
}
} else {
- wfSuppressWarnings();
+ MediaWiki\suppressWarnings();
array_push( $tagstack, $ot );
- wfRestoreWarnings();
+ MediaWiki\restoreWarnings();
# <li> can be nested in <ul> or <ol>, skip those cases:
if ( !isset( $htmllist[$ot] ) || !isset( $listtags[$t] ) ) {
# Double-quoted
return $set[3];
} elseif ( !isset( $set[2] ) ) {
- # In XHTML, attributes must have a value.
- # For 'reduced' form, return explicitly the attribute name here.
- return $set[1];
+ # In XHTML, attributes must have a value so return an empty string.
+ # See "Empty attribute syntax",
+ # http://www.w3.org/TR/html5/syntax.html#syntax-attribute-name
+ return "";
} else {
throw new MWException( "Tag conditions not met. This should never happen and is a bug." );
}