Thijs/wiki.techinc.nl
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

Walk the dom instead of using a queryselector in cleanUpTocLine

Browse source 
Change-Id: Ic59a4883f5b830c0c513e1836ad0de7c29a4b96d
This commit is contained in:
Arlo Breault 2024-02-20 17:49:15 -05:00
parent 89ddae6805
commit b05e4b98ce
1 changed files with 36 additions and 27 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

63
includes/parser/Parser.php
View file

@ -102,6 +102,8 @@ use Wikimedia\IPUtils;
use Wikimedia\Parsoid\Core\SectionMetadata;
use Wikimedia\Parsoid\Core\TOCData;
use Wikimedia\Parsoid\DOM\DocumentFragment;
use Wikimedia\Parsoid\DOM\Element;
use Wikimedia\Parsoid\DOM\Node;
use Wikimedia\Parsoid\Utils\DOMCompat;
use Wikimedia\Parsoid\Utils\DOMUtils;
use Wikimedia\ScopedCallback;
@ -4142,7 +4144,8 @@ class Parser {
->page( $this->getPage() );
}
private function cleanUpTocLine( DocumentFragment $container ) {
private function cleanUpTocLine( Node $container ) {
'@phan-var Element|DocumentFragment $container'; // @var Element|DocumentFragment $container
# Strip out HTML
# Allowed tags are:
# * <sup> and <sub> (T10393)
@ -4155,35 +4158,41 @@ class Parser {
# We strip any parameter from accepted tags, except dir="rtl|ltr" from <span>,
# to allow setting directionality in toc items.
$allowedTags = [ 'span', 'sup', 'sub', 'bdi', 'i', 'b', 's', 'strike', 'q' ];
$allNodes = DOMCompat::querySelectorAll( $container, '*' );
foreach ( $allNodes as $node ) {
$nodeName = DOMCompat::nodeName( $node );
if ( in_array( $nodeName, [ 'style', 'script' ], true ) ) {
# Remove any <style> or <script> tags (T198618)
DOMCompat::remove( $node );
} elseif ( in_array( $nodeName, $allowedTags, true ) ) {
// Keep tag, remove attributes
$removeAttrs = [];
foreach ( $node->attributes as $attr ) {
if (
$nodeName === 'span' && $attr->name === 'dir'
&& ( $attr->value === 'rtl' || $attr->value === 'ltr' )
) {
// Keep <span dir="rtl"> and <span dir="ltr">
continue;
$node = $container->firstChild;
while ( $node !== null ) {
$next = $node->nextSibling;
if ( $node instanceof Element ) {
$nodeName = DOMCompat::nodeName( $node );
if ( in_array( $nodeName, [ 'style', 'script' ], true ) ) {
# Remove any <style> or <script> tags (T198618)
DOMCompat::remove( $node );
} elseif ( in_array( $nodeName, $allowedTags, true ) ) {
// Keep tag, remove attributes
$removeAttrs = [];
foreach ( $node->attributes as $attr ) {
if (
$nodeName === 'span' && $attr->name === 'dir'
&& ( $attr->value === 'rtl' || $attr->value === 'ltr' )
) {
// Keep <span dir="rtl"> and <span dir="ltr">
continue;
}
$removeAttrs[] = $attr;
}
$removeAttrs[] = $attr;
foreach ( $removeAttrs as $attr ) {
$node->removeAttributeNode( $attr );
}
$this->cleanUpTocLine( $node );
} else {
// Strip tag
$next = $node->firstChild;
while ( $childNode = $node->firstChild ) {
$node->parentNode->insertBefore( $childNode, $node );
}
DOMCompat::remove( $node );
}
foreach ( $removeAttrs as $attr ) {
$node->removeAttributeNode( $attr );
}
} else {
// Strip tag
while ( $childNode = $node->firstChild ) {
$node->parentNode->insertBefore( $childNode, $node );
}
DOMCompat::remove( $node );
}
$node = $next;
}
# Strip '<span></span>', which is the result from the above if