Fix Rich Media Previews for updated activities

The Rich Media Previews were not regenerated when a post was updated due to a cache invalidation issue. They are now cached by the activity id so they can be evicted with the other activity cache objects in the :scrubber_cache.

3 files changed, 32 insertions, 22 deletions

diff --git a/lib/pleroma/activity/html.ex b/lib/pleroma/activity/html.ex
index 706b2d36c..ba284b4d5 100644
--- a/lib/pleroma/activity/html.ex
+++ b/lib/pleroma/activity/html.ex
@@ -28,7 +28,7 @@ defmodule Pleroma.Activity.HTML do
     end
   end
 
-  defp add_cache_key_for(activity_id, additional_key) do
+  def add_cache_key_for(activity_id, additional_key) do
     current = get_cache_keys_for(activity_id)
 
     unless additional_key in current do
diff --git a/lib/pleroma/html.ex b/lib/pleroma/html.ex
index 5bf735c4f..84ff2f129 100644
--- a/lib/pleroma/html.ex
+++ b/lib/pleroma/html.ex
@@ -6,8 +6,6 @@ defmodule Pleroma.HTML do
   # Scrubbers are compiled on boot so they can be configured in OTP releases
   #  @on_load :compile_scrubbers
 
-  @cachex Pleroma.Config.get([:cachex, :provider], Cachex)
-
   def compile_scrubbers do
     dir = Path.join(:code.priv_dir(:pleroma), "scrubbers")
 
@@ -67,27 +65,20 @@ defmodule Pleroma.HTML do
     end
   end
 
-  def extract_first_external_url_from_object(%{data: %{"content" => content}} = object)
+  @spec extract_first_external_url_from_object(Pleroma.Object.t()) ::
+          {:ok, String.t()} | {:error, :no_content}
+  def extract_first_external_url_from_object(%{data: %{"content" => content}})
       when is_binary(content) do
-    unless object.data["fake"] do
-      key = "URL|#{object.id}"
+    url =
+      content
+      |> Floki.parse_fragment!()
+      |> Floki.find("a:not(.mention,.hashtag,.attachment,[rel~=\"tag\"])")
+      |> Enum.take(1)
+      |> Floki.attribute("href")
+      |> Enum.at(0)
 
-      @cachex.fetch!(:scrubber_cache, key, fn _key ->
-        {:commit, {:ok, extract_first_external_url(content)}}
-      end)
-    else
-      {:ok, extract_first_external_url(content)}
-    end
+    {:ok, url}
   end
 
   def extract_first_external_url_from_object(_), do: {:error, :no_content}
-
-  def extract_first_external_url(content) do
-    content
-    |> Floki.parse_fragment!()
-    |> Floki.find("a:not(.mention,.hashtag,.attachment,[rel~=\"tag\"])")
-    |> Enum.take(1)
-    |> Floki.attribute("href")
-    |> Enum.at(0)
-  end
 end
diff --git a/lib/pleroma/web/rich_media/helpers.ex b/lib/pleroma/web/rich_media/helpers.ex
index 61000bb9b..ee12a9dfb 100644
--- a/lib/pleroma/web/rich_media/helpers.ex
+++ b/lib/pleroma/web/rich_media/helpers.ex
@@ -8,6 +8,8 @@ defmodule Pleroma.Web.RichMedia.Helpers do
   alias Pleroma.Object
   alias Pleroma.Web.RichMedia.Parser
 
+  @cachex Pleroma.Config.get([:cachex, :provider], Cachex)
+
   @config_impl Application.compile_env(:pleroma, [__MODULE__, :config_impl], Pleroma.Config)
 
   @options [
@@ -71,7 +73,24 @@ defmodule Pleroma.Web.RichMedia.Helpers do
   def fetch_data_for_activity(%Activity{data: %{"type" => "Create"}} = activity) do
     with true <- @config_impl.get([:rich_media, :enabled]),
          %Object{} = object <- Object.normalize(activity, fetch: false) do
-      fetch_data_for_object(object)
+      if object.data["fake"] do
+        fetch_data_for_object(object)
+      else
+        key = "URL|#{activity.id}"
+
+        @cachex.fetch!(:scrubber_cache, key, fn _ ->
+          result = fetch_data_for_object(object)
+
+          cond do
+            match?(%{page_url: _, rich_media: _}, result) ->
+              Activity.HTML.add_cache_key_for(activity.id, key)
+              {:commit, result}
+
+            true ->
+              {:ignore, %{}}
+          end
+        end)
+      end
     else
       _ -> %{}
     end