Merge branch 'fix/extract_first_external_url' into 'develop'

Fix Pleroma.HTML.extract_first_external_url/2

Closes #1451

See merge request pleroma/pleroma!2024

lib/pleroma/html.ex

@@ -91,6 +91,7 @@ def extract_first_external_url(object, content) do
     Cachex.fetch!(:scrubber_cache, key, fn _key ->
       result =
         content
+        |> HtmlEntities.decode()
         |> Floki.filter_out("a.mention,a.hashtag,a[rel~=\"tag\"]")
         |> Floki.attribute("a", "href")
         |> Enum.at(0)

test/html_test.exs

@@ -228,5 +228,16 @@ test "skips microformats hashtags" do
 
       assert url == "https://www.pixiv.net/member_illust.php?mode=medium&illust_id=72255140"
     end
+
+    test "does not crash when there is an HTML entity in a link" do
+      user = insert(:user)
+
+      {:ok, activity} =
+        CommonAPI.post(user, %{"status" => "\"http://cofe.com/?boomer=ok&foo=bar\""})
+
+      object = Object.normalize(activity)
+
+      assert {:ok, nil} = HTML.extract_first_external_url(object, object.data["content"])
+    end
   end
 end