diff --git a/app/models/topic_embed.rb b/app/models/topic_embed.rb
index 99bc6279cef..8e45c1d60ea 100644
--- a/app/models/topic_embed.rb
+++ b/app/models/topic_embed.rb
@@ -69,7 +69,7 @@ class TopicEmbed < ActiveRecord::Base
def self.find_remote(url)
require 'ruby-readability'
- original_uri = URI.parse(url)
+ original_uri = URI.parse(URI.encode(url))
opts = {
tags: %w[div p code pre h1 h2 h3 b em i strong a img ul li ol blockquote],
attributes: %w[href src class],
@@ -81,7 +81,7 @@ class TopicEmbed < ActiveRecord::Base
embed_classname_whitelist = SiteSetting.embed_classname_whitelist if SiteSetting.embed_classname_whitelist.present?
response = FetchResponse.new
- html = open(url, allow_redirections: :safe).read
+ html = open(URI.encode(url), allow_redirections: :safe).read
raw_doc = Nokogiri::HTML(html)
auth_element = raw_doc.at('meta[@name="author"]')
@@ -107,7 +107,7 @@ class TopicEmbed < ActiveRecord::Base
src = node[url_param]
unless (src.nil? || src.empty?)
begin
- uri = URI.parse(src)
+ uri = URI.parse(URI.encode(src))
unless uri.host
uri.scheme = original_uri.scheme
uri.host = original_uri.host
@@ -145,7 +145,7 @@ class TopicEmbed < ActiveRecord::Base
# Convert any relative URLs to absolute. RSS is annoying for this.
def self.absolutize_urls(url, contents)
url = normalize_url(url)
- uri = URI(url)
+ uri = URI(URI.encode(url))
prefix = "#{uri.scheme}://#{uri.host}"
prefix << ":#{uri.port}" if uri.port != 80 && uri.port != 443
diff --git a/lib/topic_retriever.rb b/lib/topic_retriever.rb
index 7401619c36f..672ca60a696 100644
--- a/lib/topic_retriever.rb
+++ b/lib/topic_retriever.rb
@@ -34,6 +34,7 @@ class TopicRetriever
# It's possible another process or job found the embed already. So if that happened bail out.
return if TopicEmbed.where(embed_url: @embed_url).exists?
+
# First check RSS if that is enabled
if SiteSetting.feed_polling_enabled?
Jobs::PollFeed.new.execute({})
diff --git a/spec/models/topic_embed_spec.rb b/spec/models/topic_embed_spec.rb
index c7bfa0d43a3..7270c4fc7e9 100644
--- a/spec/models/topic_embed_spec.rb
+++ b/spec/models/topic_embed_spec.rb
@@ -176,7 +176,23 @@ describe TopicEmbed do
it 'img node doesn\'t have other class' do
expect(response.body).to have_tag('img', without: { class: 'other' })
end
+ end
+ context "non-ascii URL" do
+ let(:url) { 'http://eviltrout.com/test/ماهی' }
+ let(:contents) { "<title>سلام</title><body>این یک پاراگراف آزمون است.</body>" }
+ let!(:embeddable_host) { Fabricate(:embeddable_host) }
+ let!(:file) { StringIO.new }
+
+ before do
+ file.stubs(:read).returns contents
+ TopicEmbed.stubs(:open).returns file
+ end
+
+ it "doesn't throw an error" do
+ response = TopicEmbed.find_remote(url)
+ expect(response.title).to eq("سلام")
+ end
end
end