Merge branch 'feat/force-mentions-mrf' into 'develop'

Add ForceMentionsInContentPolicy

See merge request pleroma/pleroma!3609

CHANGELOG.md

@@ -16,6 +16,7 @@ The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/).
 ### Added
 - `activeMonth` and `activeHalfyear` fields in NodeInfo usage.users object
 - Experimental support for Finch. Put `config :tesla, :adapter, {Tesla.Adapter.Finch, name: MyFinch}` in your secrets file to use it. Reverse Proxy will still use Hackney.
+- `ForceMentionsInPostContent` MRF policy
 - AdminAPI: allow moderators to manage reports, users, invites, and custom emojis
 - AdminAPI: restrict moderators to access sensitive data: change user credentials, get password reset token, read private statuses and chats, etc
 - PleromaAPI: Add remote follow API endpoint at `POST /api/v1/pleroma/remote_interaction`

docs/configuration/cheatsheet.md

@@ -126,6 +126,7 @@ To add configuration to your config file, you can copy it from the base config.
     * `Pleroma.Web.ActivityPub.MRF.ForceBotUnlistedPolicy`: Makes all bot posts to disappear from public timelines.
     * `Pleroma.Web.ActivityPub.MRF.FollowBotPolicy`: Automatically follows newly discovered users from the specified bot account. Local accounts, locked accounts, and users with "#nobot" in their bio are respected and excluded from being followed.
     * `Pleroma.Web.ActivityPub.MRF.KeywordPolicy`: Rejects or removes from the federated timeline or replaces keywords. (See [`:mrf_keyword`](#mrf_keyword)).
+    * `Pleroma.Web.ActivityPub.MRF.ForceMentionsInContent`: Forces every mentioned user to be reflected in the post content.
 * `transparency`: Make the content of your Message Rewrite Facility settings public (via nodeinfo).
 * `transparency_exclusions`: Exclude specific instance names from MRF transparency.  The use of the exclusions feature will be disclosed in nodeinfo as a boolean value.
 

lib/pleroma/formatter.ex

@@ -34,11 +34,19 @@ def escape_mention_handler("@" <> nickname = mention, buffer, _, _) do
 
   def mention_handler("@" <> nickname, buffer, opts, acc) do
     case User.get_cached_by_nickname(nickname) do
-      %User{id: id} = user ->
-        user_url = user.uri || user.ap_id
-        nickname_text = get_nickname_text(nickname, opts)
+      %User{} = user ->
+        {mention_from_user(user, opts),
+         %{acc | mentions: MapSet.put(acc.mentions, {"@" <> nickname, user})}}
+
+      _ ->
+        {buffer, acc}
+    end
+  end
+
+  def mention_from_user(%User{id: id} = user, opts \\ %{mentions_format: :full}) do
+    user_url = user.uri || user.ap_id
+    nickname_text = get_nickname_text(user.nickname, opts)
 
-        link =
     Phoenix.HTML.Tag.content_tag(
       :span,
       Phoenix.HTML.Tag.content_tag(
@@ -52,12 +60,6 @@ def mention_handler("@" <> nickname, buffer, opts, acc) do
       class: "h-card"
     )
     |> Phoenix.HTML.safe_to_string()
-
-        {link, %{acc | mentions: MapSet.put(acc.mentions, {"@" <> nickname, user})}}
-
-      _ ->
-        {buffer, acc}
-    end
   end
 
   def hashtag_handler("#" <> tag = tag_text, _buffer, _opts, acc) do

lib/pleroma/user.ex

@@ -1055,6 +1055,10 @@ def get_by_ap_id(ap_id) do
     Repo.get_by(User, ap_id: ap_id)
   end
 
+  def get_by_uri(uri) do
+    Repo.get_by(User, uri: uri)
+  end
+
   def get_all_by_ap_id(ap_ids) do
     from(u in __MODULE__,
       where: u.ap_id in ^ap_ids

lib/pleroma/web/activity_pub/mrf/force_mentions_in_content.ex

@@ -0,0 +1,80 @@
+# Pleroma: A lightweight social networking server
+# Copyright © 2017-2022 Pleroma Authors <https://pleroma.social/>
+# SPDX-License-Identifier: AGPL-3.0-only
+
+defmodule Pleroma.Web.ActivityPub.MRF.ForceMentionsInContent do
+  alias Pleroma.Formatter
+  alias Pleroma.User
+
+  @behaviour Pleroma.Web.ActivityPub.MRF.Policy
+
+  defp do_extract({:a, attrs, _}, acc) do
+    if Enum.find(attrs, fn {name, value} ->
+         name == "class" && value in ["mention", "u-url mention", "mention u-url"]
+       end) do
+      href = Enum.find(attrs, fn {name, _} -> name == "href" end) |> elem(1)
+      acc ++ [href]
+    else
+      acc
+    end
+  end
+
+  defp do_extract({_, _, children}, acc) do
+    do_extract(children, acc)
+  end
+
+  defp do_extract(nodes, acc) when is_list(nodes) do
+    Enum.reduce(nodes, acc, fn node, acc -> do_extract(node, acc) end)
+  end
+
+  defp do_extract(_, acc), do: acc
+
+  defp extract_mention_uris_from_content(content) do
+    {:ok, tree} = :fast_html.decode(content, format: [:html_atoms])
+    do_extract(tree, [])
+  end
+
+  @impl true
+  def filter(%{"type" => "Create", "object" => %{"type" => "Note", "tag" => tag}} = object) do
+    # image-only posts from pleroma apparently reach this MRF without the content field
+    content = object["object"]["content"] || ""
+
+    mention_users =
+      tag
+      |> Enum.filter(fn tag -> tag["type"] == "Mention" end)
+      |> Enum.map(& &1["href"])
+      |> Enum.reject(&is_nil/1)
+      |> Enum.map(fn ap_id_or_uri ->
+        case User.get_or_fetch_by_ap_id(ap_id_or_uri) do
+          {:ok, user} -> {ap_id_or_uri, user}
+          _ -> {ap_id_or_uri, User.get_by_uri(ap_id_or_uri)}
+        end
+      end)
+      |> Enum.reject(fn {_, user} -> user == nil end)
+      |> Enum.into(%{})
+
+    explicitly_mentioned_uris = extract_mention_uris_from_content(content)
+
+    added_mentions =
+      Enum.reduce(mention_users, "", fn {uri, user}, acc ->
+        unless uri in explicitly_mentioned_uris do
+          acc <> Formatter.mention_from_user(user)
+        else
+          acc
+        end
+      end)
+
+    content =
+      if added_mentions != "",
+        do: added_mentions <> " " <> content,
+        else: content
+
+    {:ok, put_in(object["object"]["content"], content)}
+  end
+
+  @impl true
+  def filter(object), do: {:ok, object}
+
+  @impl true
+  def describe, do: {:ok, %{}}
+end

test/fixtures/soapbox_no_mentions_in_content.json

@@ -0,0 +1,79 @@
+{
+  "@context": [
+    "https://www.w3.org/ns/activitystreams",
+    "https://gleasonator.com/schemas/litepub-0.1.jsonld",
+    {
+      "@language": "und"
+    }
+  ],
+  "actor": "https://gleasonator.com/users/alex",
+  "attachment": [
+    {
+      "blurhash": "b15#-6_3~l%eDkNraAM#HYMf",
+      "height": 2147,
+      "mediaType": "image/png",
+      "name": "",
+      "type": "Document",
+      "url": "https://media.gleasonator.com/2df1c9cc26c35028db65848143971da0c5a6e04dbd423b194bb018b6c601db9b.png",
+      "width": 966
+    },
+    {
+      "blurhash": "b168EX~q~W-;DiM{VtIUD%Io",
+      "height": 2147,
+      "mediaType": "image/png",
+      "name": "",
+      "type": "Document",
+      "url": "https://media.gleasonator.com/22b42b4cddc1aecc7e2d3dc20bcdd66d5c4e44c2bef119de948e95a587e36b66.png",
+      "width": 966
+    }
+  ],
+  "attributedTo": "https://gleasonator.com/users/alex",
+  "cc": [
+    "https://www.w3.org/ns/activitystreams#Public"
+  ],
+  "content": "<p>Haha yeah, you can control who you reply to.</p>",
+  "context": "https://gleasonator.com/contexts/ba6c8bd9-ac4d-479e-9bd9-5bf570068ae7",
+  "conversation": "https://gleasonator.com/contexts/ba6c8bd9-ac4d-479e-9bd9-5bf570068ae7",
+  "id": "https://gleasonator.com/objects/02af65fe-04f8-46bc-9b1e-31dfe76eaffd",
+  "inReplyTo": "https://shitposter.club/objects/c686d811-4368-48e1-ba11-82c129f93165",
+  "published": "2022-01-19T03:37:35.976545Z",
+  "sensitive": false,
+  "source": "Haha yeah, you can control who you reply to.",
+  "summary": "",
+  "tag": [
+    {
+      "href": "https://lain.com/users/lain",
+      "name": "@lain@lain.com",
+      "type": "Mention"
+    },
+    {
+      "href": "https://shitposter.club/users/coolboymew",
+      "name": "@coolboymew@shitposter.club",
+      "type": "Mention"
+    },
+    {
+      "href": "https://shitposter.club/users/dielan",
+      "name": "@dielan@shitposter.club",
+      "type": "Mention"
+    },
+    {
+      "href": "https://tuusin.misono-ya.info/users/hakui",
+      "name": "@hakui@tuusin.misono-ya.info",
+      "type": "Mention"
+    },
+    {
+      "href": "https://xyzzy.link/users/fence",
+      "name": "@fence@xyzzy.link",
+      "type": "Mention"
+    }
+  ],
+  "to": [
+    "https://shitposter.club/users/dielan",
+    "https://gleasonator.com/users/alex/followers",
+    "https://shitposter.club/users/coolboymew",
+    "https://xyzzy.link/users/fence",
+    "https://tuusin.misono-ya.info/users/hakui",
+    "https://lain.com/users/lain"
+  ],
+  "type": "Note"
+}

test/pleroma/web/activity_pub/mrf/force_mentions_in_content_test.exs

@@ -0,0 +1,34 @@
+# Pleroma: A lightweight social networking server
+# Copyright © 2017-2021 Pleroma Authors <https://pleroma.social/>
+# SPDX-License-Identifier: AGPL-3.0-only
+
+defmodule Pleroma.Web.ActivityPub.MRF.ForceMentionsInContentTest do
+  alias Pleroma.Web.ActivityPub.MRF.ForceMentionsInContent
+  import Pleroma.Factory
+  use Pleroma.DataCase
+
+  test "adds mentions to post content" do
+    users = %{
+      "lain@lain.com" => "https://lain.com/users/lain",
+      "coolboymew@shitposter.club" => "https://shitposter.club/users/coolboymew",
+      "dielan@shitposter.club" => "https://shitposter.club/users/dielan",
+      "hakui@tuusin.misono-ya.info" => "https://tuusin.misono-ya.info/users/hakui",
+      "fence@xyzzy.link" => "https://xyzzy.link/users/fence"
+    }
+
+    Enum.each(users, fn {nickname, ap_id} ->
+      insert(:user, ap_id: ap_id, nickname: nickname, local: false)
+    end)
+
+    object = File.read!("test/fixtures/soapbox_no_mentions_in_content.json") |> Jason.decode!()
+
+    activity = %{
+      "type" => "Create",
+      "actor" => "https://gleasonator.com/users/alex",
+      "object" => object
+    }
+
+    {:ok, %{"object" => %{"content" => filtered}}} = ForceMentionsInContent.filter(activity)
+    Enum.each(users, fn {nickname, _} -> assert filtered =~ nickname end)
+  end
+end