diff --git a/app/models/topic_embed.rb b/app/models/topic_embed.rb
index 39cab1a7983..bb32d3ab4db 100644
--- a/app/models/topic_embed.rb
+++ b/app/models/topic_embed.rb
@@ -114,17 +114,27 @@ class TopicEmbed < ActiveRecord::Base
end
def self.find_remote(url)
- require "ruby-readability"
-
url = UrlHelper.normalized_encode(url)
- original_uri = URI.parse(url)
+ URI.parse(url) # ensure url parses, will raise if not
fd = FinalDestination.new(url, validate_uri: true, max_redirects: 5, follow_canonical: true)
uri = fd.resolve
return if uri.blank?
+ begin
+ html = uri.read
+ rescue OpenURI::HTTPError, Net::OpenTimeout
+ return
+ end
+
+ parse_html(html, url)
+ end
+
+ def self.parse_html(html, url)
+ require "ruby-readability"
+
opts = {
- tags: %w[div p code pre h1 h2 h3 b em i strong a img ul li ol blockquote],
+ tags: %w[div p code pre h1 h2 h3 b em i strong a img ul li ol blockquote figure figcaption],
attributes: %w[href src class],
remove_empty_nodes: false,
}
@@ -139,11 +149,6 @@ class TopicEmbed < ActiveRecord::Base
SiteSetting.allowed_embed_classnames if SiteSetting.allowed_embed_classnames.present?
response = FetchResponse.new
- begin
- html = uri.read
- rescue OpenURI::HTTPError, Net::OpenTimeout
- return
- end
raw_doc = Nokogiri.HTML5(html)
auth_element =
@@ -200,7 +205,7 @@ class TopicEmbed < ActiveRecord::Base
end
end
- response.body = doc.to_html
+ response.body = doc.at("body").children.to_html
response
end
diff --git a/spec/models/topic_embed_spec.rb b/spec/models/topic_embed_spec.rb
index 82ed4957e66..ede62d5c7c4 100644
--- a/spec/models/topic_embed_spec.rb
+++ b/spec/models/topic_embed_spec.rb
@@ -23,6 +23,41 @@ RSpec.describe TopicEmbed do
expect(TopicEmbed.count).to eq(0)
end
+ it "Allows figure and figcaption HTML tags" do
+ html = <<~HTML
+
+
+ Some title
+
+
+
+
some content
+
+
+ Some caption
+
+
+
+
+ HTML
+
+ parsed = TopicEmbed.parse_html(html, "https://blog.discourse.com/somepost.html")
+
+ # div inception is inserted by the readability gem
+ expected = <<~HTML
+
+
+
some content
+
+
+ Some caption
+
+
+
+ HTML
+ expect(parsed.body.strip).to eq(expected.strip)
+ end
+
context "when creating a post" do
let!(:post) { TopicEmbed.import(user, url, title, contents) }
let(:topic_embed) { TopicEmbed.find_by(post: post) }
+ 
+