Clarify when we are scrubbing vs when we are filtering so we can keep newlines

lib/pleroma/web/metadata/providers/open_graph.ex

@@ -19,7 +19,7 @@ defmodule Pleroma.Web.Metadata.Providers.OpenGraph do
         user: user
       }) do
     attachments = build_attachments(object)
-    scrubbed_content = Utils.scrub_html_and_truncate(object)
+    filtered_content = Utils.filter_html_and_truncate(object)
 
     [
       {:meta,
@@ -31,7 +31,7 @@ defmodule Pleroma.Web.Metadata.Providers.OpenGraph do
       {:meta,
        [
          property: "og:description",
-         content: scrubbed_content
+         content: filtered_content
        ], []},
       {:meta, [property: "og:type", content: "article"], []}
     ] ++
@@ -49,7 +49,7 @@ defmodule Pleroma.Web.Metadata.Providers.OpenGraph do
 
   @impl Provider
   def build_tags(%{user: user}) do
-    with truncated_bio = Utils.scrub_html_and_truncate(user.bio) do
+    with truncated_bio = Utils.filter_html_and_truncate(user.bio) do
       [
         {:meta,
          [

lib/pleroma/web/metadata/providers/twitter_card.ex

@@ -16,11 +16,11 @@ defmodule Pleroma.Web.Metadata.Providers.TwitterCard do
   @impl Provider
   def build_tags(%{activity_id: id, object: object, user: user}) do
     attachments = build_attachments(id, object)
-    scrubbed_content = Utils.scrub_html_and_truncate(object)
+    filtered_content = Utils.filter_html_and_truncate(object)
 
     [
       title_tag(user),
-      {:meta, [property: "twitter:description", content: scrubbed_content], []}
+      {:meta, [property: "twitter:description", content: filtered_content], []}
     ] ++
       if attachments == [] or Metadata.activity_nsfw?(object) do
         [
@@ -34,7 +34,7 @@ defmodule Pleroma.Web.Metadata.Providers.TwitterCard do
 
   @impl Provider
   def build_tags(%{user: user}) do
-    with truncated_bio = Utils.scrub_html_and_truncate(user.bio) do
+    with truncated_bio = Utils.filter_html_and_truncate(user.bio) do
       [
         title_tag(user),
         {:meta, [property: "twitter:description", content: truncated_bio], []},

lib/pleroma/web/metadata/utils.ex

@@ -7,25 +7,17 @@ defmodule Pleroma.Web.Metadata.Utils do
   alias Pleroma.Formatter
   alias Pleroma.HTML
 
-  def filter_html_and_truncate(%{data: %{"content" => content}} = _object) do
-    content
-    # html content comes from DB already encoded, decode first and scrub after
-    |> Emoji.Formatter.demojify()
-    |> HTML.filter_tags(Pleroma.HTML.Scrubber.BreaksOnly)
-    |> HtmlEntities.decode()
-    |> String.replace(~r/<br\s?\/?>/, "&#10;&#13;")
-    |> Formatter.truncate()
-  end
+  def filter_html_and_truncate(%{data: %{"content" => content}} = _object),
+    do: do_filter_html_and_truncate(content)
 
-  def scrub_html_and_truncate(content, max_length \\ 200) when is_binary(content) do
-    content
-    |> Emoji.Formatter.demojify()
-    |> HtmlEntities.decode()
-    |> String.replace(~r/<br\s?\/?>/, " ")
-    |> HTML.strip_tags()
-    |> HtmlEntities.decode()
-    |> Formatter.truncate(max_length)
-  end
+  def filter_html_and_truncate(content, max_length \\ nil),
+    do: do_filter_html_and_truncate(content, max_length)
+
+  def scrub_html_and_truncate(%{data: %{"content" => content}} = _object),
+    do: do_scrub_html_and_truncate(content)
+
+  def scrub_html_and_truncate(content, max_length \\ nil),
+    do: do_scrub_html_and_truncate(content, max_length)
 
   def user_name_string(user) do
     "#{user.name} " <>
@@ -42,4 +34,25 @@ defmodule Pleroma.Web.Metadata.Utils do
       String.starts_with?(media_type, support_type)
     end)
   end
+
+  defp do_filter_html_and_truncate(content, max_length \\ 200) when is_binary(content) do
+    # html content comes from DB already encoded, but demojify decodes for us
+    content
+    |> Emoji.Formatter.demojify()
+    |> HTML.filter_tags(Pleroma.HTML.Scrubber.BreaksOnly)
+    |> HtmlEntities.decode()
+    |> String.replace(~r/<br\s?\/?>/, "&#10;&#13;")
+    |> Formatter.truncate(max_length)
+  end
+
+  defp do_scrub_html_and_truncate(content, max_length \\ 200) when is_binary(content) do
+    # html content comes from DB already encoded, but demojify decodes for us
+    content
+    |> Emoji.Formatter.demojify()
+    |> HtmlEntities.decode()
+    |> String.replace(~r/<br\s?\/?>/, " ")
+    |> HTML.strip_tags()
+    |> HtmlEntities.decode()
+    |> Formatter.truncate(max_length)
+  end
 end