Mike Dalessio
95a45aa651
The scrubber was memoized on the view helper in b0fa6525 so that
@regexp_timeout_reported would stop scanning remaining text nodes
after a timeout. But memoizing on the helper leaks that flag across
all format_html calls for the entire request — if one comment triggers
a timeout, every subsequent comment on the page loses auto-linking.
Use a fresh scrubber per call. The @regexp_timeout_reported flag still
works within a single document's scrub, which is the intended scope.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
12 lines
311 B
Ruby
12 lines
311 B
Ruby
module HtmlHelper
|
|
def format_html(html)
|
|
Loofah::HTML5::DocumentFragment.parse(html).scrub!(AutoLinkScrubber.new).to_html.html_safe
|
|
end
|
|
|
|
def card_html_title(card)
|
|
return card.title if card.title.blank?
|
|
|
|
ERB::Util.html_escape(card.title).gsub(/`([^`]+)`/, '<code>\1</code>').html_safe
|
|
end
|
|
end
|