Remove tests; function is no longer used

They shouldn't share the same name when /1 was used for a different type of incoming data anyway

lib/pleroma/web/feed/feed_view.ex

@@ -7,6 +7,7 @@ defmodule Pleroma.Web.Feed.FeedView do
   use Pleroma.Web, :view
 
   alias Pleroma.Formatter
+  alias Pleroma.HTML
   alias Pleroma.Object
   alias Pleroma.User
   alias Pleroma.Web.MediaProxy
@@ -70,7 +71,9 @@ defmodule Pleroma.Web.Feed.FeedView do
 
   def activity_title(%{"content" => content}, opts \\ %{}) do
     content
-    |> Pleroma.Web.Metadata.Utils.scrub_html()
+    |> HtmlEntities.decode()
+    |> String.replace(~r/<br\s?\/?>/, " ")
+    |> HTML.strip_tags()
     |> Pleroma.Emoji.Formatter.demojify()
     |> Formatter.truncate(opts[:max_length], opts[:omission])
     |> escape()

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

@@ -3,39 +3,15 @@
 # SPDX-License-Identifier: AGPL-3.0-only
 
 defmodule Pleroma.Web.Metadata.Utils do
-  alias Pleroma.Activity
   alias Pleroma.Emoji
   alias Pleroma.Formatter
   alias Pleroma.HTML
 
-  def scrub_html_and_truncate(%{data: %{"content" => content}} = object) do
+  def filter_html_and_truncate(%{data: %{"content" => content}} = _object),
-    content
+    do: do_filter_html_and_truncate(content)
-    # html content comes from DB already encoded, decode first and scrub after
-    |> HtmlEntities.decode()
-    |> String.replace(~r/<br\s?\/?>/, " ")
-    |> Activity.HTML.get_cached_stripped_html_for_activity(object, "metadata")
-    |> Emoji.Formatter.demojify()
-    |> HtmlEntities.decode()
-    |> Formatter.truncate()
-  end
 
-  def scrub_html_and_truncate(content, max_length \\ 200) when is_binary(content) do
+  def filter_html_and_truncate(content, max_length \\ nil),
-    content
+    do: do_filter_html_and_truncate(content, max_length)
-    |> scrub_html
-    |> Emoji.Formatter.demojify()
-    |> HtmlEntities.decode()
-    |> Formatter.truncate(max_length)
-  end
-
-  def scrub_html(content) when is_binary(content) do
-    content
-    # html content comes from DB already encoded, decode first and scrub after
-    |> HtmlEntities.decode()
-    |> String.replace(~r/<br\s?\/?>/, " ")
-    |> HTML.strip_tags()
-  end
-
-  def scrub_html(content), do: content
 
   def user_name_string(user) do
     "#{user.name} " <>
@@ -52,4 +28,15 @@ 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
+    content
+    |> HtmlEntities.decode()
+    |> Emoji.Formatter.demojify()
+    |> HTML.filter_tags(Pleroma.HTML.Scrubber.BreaksOnly)
+    |> HtmlEntities.decode()
+    |> String.replace(~r/<br\s?\/?>/, "&#10;&#13;")
+    |> Formatter.truncate(max_length)
+  end
 end

lib/pleroma/web/push/impl.ex

@@ -6,11 +6,13 @@ defmodule Pleroma.Web.Push.Impl do
   @moduledoc "The module represents implementation push web notification"
 
   alias Pleroma.Activity
+  alias Pleroma.Emoji
+  alias Pleroma.Formatter
+  alias Pleroma.HTML
   alias Pleroma.Notification
   alias Pleroma.Object
   alias Pleroma.Repo
   alias Pleroma.User
-  alias Pleroma.Web.Metadata.Utils
   alias Pleroma.Web.Push.Subscription
 
   require Logger
@@ -127,7 +129,7 @@ defmodule Pleroma.Web.Push.Impl do
   def format_body(_activity, actor, %{data: %{"type" => "ChatMessage", "content" => content}}, _) do
     case content do
       nil -> "@#{actor.nickname}: (Attachment)"
-      content -> "@#{actor.nickname}: #{Utils.scrub_html_and_truncate(content, 80)}"
+      content -> "@#{actor.nickname}: #{filter_html_and_truncate(content, 80)}"
     end
   end
 
@@ -137,7 +139,7 @@ defmodule Pleroma.Web.Push.Impl do
         %{data: %{"content" => content}},
         _mastodon_type
       ) do
-    "@#{actor.nickname}: #{Utils.scrub_html_and_truncate(content, 80)}"
+    "@#{actor.nickname}: #{filter_html_and_truncate(content, 80)}"
   end
 
   def format_body(
@@ -146,7 +148,7 @@ defmodule Pleroma.Web.Push.Impl do
         %{data: %{"content" => content}},
         _mastodon_type
       ) do
-    "@#{actor.nickname} repeated: #{Utils.scrub_html_and_truncate(content, 80)}"
+    "@#{actor.nickname} repeated: #{filter_html_and_truncate(content, 80)}"
   end
 
   def format_body(
@@ -192,4 +194,15 @@ defmodule Pleroma.Web.Push.Impl do
       type -> "New #{String.capitalize(type || "event")}"
     end
   end
+
+  defp filter_html_and_truncate(content, max_length) when is_binary(content) do
+    # html content comes from DB already encoded
+    content
+    |> HtmlEntities.decode()
+    |> Emoji.Formatter.demojify()
+    |> HTML.filter_tags(Pleroma.HTML.Scrubber.BreaksOnly)
+    |> HtmlEntities.decode()
+    |> String.replace(~r/<br\s?\/?>/, "\r\n")
+    |> Formatter.truncate(max_length)
+  end
 end

priv/scrubbers/breaks_only.ex

@@ -0,0 +1,15 @@
+defmodule Pleroma.HTML.Scrubber.BreaksOnly do
+  @moduledoc """
+  An HTML scrubbing policy which limits to linebreaks only.
+  """
+
+  require FastSanitize.Sanitizer.Meta
+  alias FastSanitize.Sanitizer.Meta
+
+  Meta.strip_comments()
+
+  # linebreaks only
+  Meta.allow_tag_with_these_attributes(:br, [])
+
+  Meta.strip_everything_not_covered()
+end

test/pleroma/web/metadata/utils_test.exs

@@ -7,8 +7,8 @@ defmodule Pleroma.Web.Metadata.UtilsTest do
   import Pleroma.Factory
   alias Pleroma.Web.Metadata.Utils
 
-  describe "scrub_html_and_truncate/1" do
+  describe "filter_html_and_truncate/1" do
-    test "it returns text without encode HTML" do
+    test "it returns text without encoded HTML entities" do
       user = insert(:user)
 
       note =
@@ -20,13 +20,55 @@ defmodule Pleroma.Web.Metadata.UtilsTest do
           }
         })
 
-      assert Utils.scrub_html_and_truncate(note) == "Pleroma's really cool!"
+      assert Utils.filter_html_and_truncate(note) == "Pleroma's really cool!"
     end
-  end
 
-  describe "scrub_html_and_truncate/2" do
+    test "it replaces <br> with compatible HTML entity (meta tags, push notifications)" do
-    test "it returns text without encode HTML" do
+      user = insert(:user)
-      assert Utils.scrub_html_and_truncate("Pleroma's really cool!") == "Pleroma's really cool!"
+
+      note =
+        insert(:note, %{
+          data: %{
+            "actor" => user.ap_id,
+            "id" => "https://pleroma.gov/objects/whatever",
+            "content" => "First line<br>Second line"
+          }
+        })
+
+      assert Utils.filter_html_and_truncate(note) ==
+               "First line&#10;&#13;Second line"
+    end
+
+    test "it strips emojis" do
+      user = insert(:user)
+
+      note =
+        insert(:note, %{
+          data: %{
+            "actor" => user.ap_id,
+            "id" => "https://pleroma.gov/objects/whatever",
+            "content" => "Mozilla Firefox :firefox:"
+          }
+        })
+
+      assert Utils.filter_html_and_truncate(note) ==
+               "Mozilla Firefox"
+    end
+
+    test "it strips HTML tags and other entities" do
+      user = insert(:user)
+
+      note =
+        insert(:note, %{
+          data: %{
+            "actor" => user.ap_id,
+            "id" => "https://pleroma.gov/objects/whatever",
+            "content" => "<title>my title</title> <p>and a paragraph&#33;</p>"
+          }
+        })
+
+      assert Utils.filter_html_and_truncate(note) ==
+               "my title and a paragraph!"
     end
   end
 end

test/pleroma/web/push/impl_test.exs

@@ -359,4 +359,29 @@ defmodule Pleroma.Web.Push.ImplTest do
              }
     end
   end
+
+  test "body for create activity handles newlines" do
+    user = insert(:user, nickname: "bob")
+    _user2 = insert(:user, nickname: "alice")
+
+    {:ok, activity} =
+      CommonAPI.post(user, %{
+        status: """
+        @alice Line one
+        Line two
+        Line three
+        """
+      })
+
+    object = Object.normalize(activity, fetch: false)
+
+    assert Impl.format_body(
+             %{
+               activity: activity
+             },
+             user,
+             object
+           ) ==
+             "@bob: @alice Line one\r\nLine two\r\nLine three"
+  end
 end