Add Article support on a per-category basis

An Article is for when you don't want to restrict the length of the content being federated (i.e. you want to federate entire posts). Note that Mastodon currently converts the content of Article types into a link, however platforms like Lemmy will show the full content. See further mastodon/mastodon#24079

app/models/concerns/discourse_activity_pub/ap/identifier_validations.rb

@@ -27,12 +27,7 @@ def _model
       end
 
       def ensure_ap_type
-        if !self.ap_type
-          self.ap_type = case _model.class.name
-                         when 'Category' then AP::Actor::Group.type
-                         when 'Post' then AP::Object::Note.type
-                         end
-        end
+        self.ap_type = _model.activity_pub_default_object_type if !self.ap_type
 
         unless ap
           self.errors.add(
@@ -42,6 +37,8 @@ def ensure_ap_type
 
           raise ActiveRecord::RecordInvalid
         end
+
+        self.ap_type = ap.type
       end
 
       def ensure_ap_key

app/serializers/discourse_activity_pub/ap/object/article_serializer.rb

@@ -0,0 +1,23 @@
+# frozen_string_literal: true
+
+class DiscourseActivityPub::AP::Object::ArticleSerializer < DiscourseActivityPub::AP::ObjectSerializer
+  attributes :content,
+             :url,
+             :updated
+
+  def include_content?
+    object.content.present? && !deleted?
+  end
+
+  def include_url?
+    object.stored.local? && !deleted?
+  end
+
+  def include_updated?
+    object.updated.present?
+  end
+
+  def deleted?
+    !object.stored.model || object.stored.model.trashed?
+  end
+end

assets/javascripts/discourse/components/activity-pub-post-object-type-dropdown.js

@@ -0,0 +1,32 @@
+import DropdownSelectBoxComponent from "select-kit/components/dropdown-select-box";
+import I18n from "I18n";
+import { computed } from "@ember/object";
+
+export default DropdownSelectBoxComponent.extend({
+  classNames: ["activity-pub-post-object-type-dropdown"],
+
+  content: computed(function () {
+    return [
+      {
+        id: "Note",
+        label: I18n.t("discourse_activity_pub.post_object_type.note.label"),
+        title: I18n.t(
+          "discourse_activity_pub.post_object_type.note.description"
+        ),
+      },
+      {
+        id: "Article",
+        label: I18n.t("discourse_activity_pub.post_object_type.article.label"),
+        title: I18n.t(
+          "discourse_activity_pub.post_object_type.article.description"
+        ),
+      },
+    ];
+  }),
+
+  actions: {
+    onChange(value) {
+      this.attrs.onChange && this.attrs.onChange(value);
+    },
+  },
+});

assets/javascripts/discourse/connectors/category-custom-settings/activity-pub-category-settings.hbs

@@ -72,4 +72,15 @@
   <div class="activity-pub-setting-description">
     {{i18n 'category.discourse_activity_pub.default_visibility_description'}}
   </div>
+</section>
+
+<section class='field activity-pub-setting activity-pub-post-object-type'>
+  <label for="activity-pub-post-object-type">{{i18n 'category.discourse_activity_pub.post_object_type'}}</label>
+  <ActivityPubPostObjectTypeDropdown
+    @value={{this.category.custom_fields.activity_pub_post_object_type}}
+    @onChange={{action (mut this.category.custom_fields.activity_pub_post_object_type)}}
+  />
+  <div class="activity-pub-setting-description">
+    {{i18n 'category.discourse_activity_pub.post_object_type_description'}}
+  </div>
 </section>

assets/javascripts/discourse/connectors/category-custom-settings/activity-pub-category-settings.js

@@ -1,5 +1,6 @@
 const customFieldDefaults = {
   activity_pub_default_visibility: "public",
+  activity_pub_post_object_type: "note",
 };
 
 export default {

assets/stylesheets/common/common.scss

@@ -132,7 +132,8 @@
   }
 }
 
-.activity-pub-visibility-dropdown {
+.activity-pub-visibility-dropdown,
+.activity-pub-post-object-type-dropdown {
   .select-kit-header {
     padding: 4px 6px;
   }

config/locales/client.en.yml

@@ -17,6 +17,8 @@ en:
         name_description: Used as a display name on some services (e.g. Mastodon).
         default_visibility: Default visibility for posts published via ActivityPub in this category.
         default_visibility_description: If the ActivityPub status is not shown in composer, all posts will be published via ActivityPub with this visibility.
+        post_object_type: ActivityPub type to publish posts in this category as.
+        post_object_type_description: Use Note for short content (e.g. Mastodon) and Article for long content.
     discourse_activity_pub:
       status:
         title:
@@ -36,10 +38,17 @@ en:
       visibility:
         private:
           label: Followers Only
-          description: ActivityPub Note is addressed to followers
+          description: ActivityPub object is addressed to followers
         public:
           label: Public
-          description: ActivityPub Note is publicly addressed
+          description: ActivityPub object is publicly addressed
+      post_object_type:
+        note:
+          label: Note
+          description: Best for microblogging platforms (e.g. Mastodon).
+        article:
+          label: Article
+          description: Best for longform platforms (e.g. Lemmy).
     post:
       discourse_activity_pub:
         title:

lib/discourse_activity_pub/ap/actor/group.rb

@@ -16,9 +16,9 @@ def can_perform_activity
           {
             accept: [:follow],
             reject: [:follow],
-            create: [:note],
-            delete: [:note],
-            update: [:note]
+            create: [:note, :article],
+            delete: [:note, :article],
+            update: [:note, :article]
           }
         end
       end

lib/discourse_activity_pub/ap/object.rb

@@ -88,7 +88,7 @@ def self.from_type(type)
 
       def self.get_klass(type)
         self.descendants.find do |klass|
-          klass.to_s.demodulize === type
+          klass.to_s.demodulize.downcase === type.downcase
         end
       end
 

lib/discourse_activity_pub/ap/object/article.rb

@@ -0,0 +1,25 @@
+# frozen_string_literal: true
+module DiscourseActivityPub
+  module AP
+    class Object
+      class Article < Object
+
+        def type
+          "Article"
+        end
+
+        def content
+          stored&.content
+        end
+
+        def updated
+          stored&.updated_at.iso8601
+        end
+
+        def can_belong_to
+          %i(post)
+        end
+      end
+    end
+  end
+end

lib/discourse_activity_pub/excerpt_parser.rb → lib/discourse_activity_pub/content_parser.rb

@@ -2,7 +2,7 @@
 
 # TODO (future): PR discourse/discourse to support alternate excerpts
 
-class DiscourseActivityPub::ExcerptParser < ExcerptParser
+class DiscourseActivityPub::ContentParser < ExcerptParser
   CUSTOM_NOTE_REGEX = /<\s*(div)[^>]*class\s*=\s*['"]note['"][^>]*>/
 
   MARKDOWN_FEATURES = %w[
@@ -46,26 +46,34 @@ def self.cook(text, opts = {})
         markdown_it_rules: MARKDOWN_IT_RULES,
       )
     )
-    doc = Nokogiri::HTML5.fragment(html)
-    scrubbed_html(doc)
+    scrubbed_html(html)
   end
 
-  def self.scrubbed_html(doc)
+  def self.scrubbed_html(html)
+    doc = Nokogiri::HTML5.fragment(html)
     scrubber = Loofah::Scrubber.new { |node| node.remove if node.name == "script" }
     loofah_fragment = Loofah.html5_fragment(doc.to_html)
     loofah_fragment.scrub!(scrubber).to_html
   end
 
   def self.get_content(post)
-    cooked = cook(post.raw, topic_id: post.topic_id, user_id: post.user_id)
-    max_length = SiteSetting.activity_pub_note_excerpt_maxlength
-    get_excerpt(cooked, max_length, post: post)
+    html = cook(post.raw, topic_id: post.topic_id, user_id: post.user_id)
+    type = post.activity_pub_object_type.downcase
+    return nil unless self.respond_to?("get_#{type}") && html
+    self.send("get_#{type}", html)
+  end
+
+  def self.get_article(html)
+    html
   end
 
-  def self.get_excerpt(html, length, options)
-    html ||= ""
-    length = html.length if html.include?("note") && CUSTOM_NOTE_REGEX === html
-    me = self.new(length, options)
+  def self.get_note(html)
+    length = if html.include?("note") && CUSTOM_NOTE_REGEX === html
+               html.length
+             else
+               SiteSetting.activity_pub_note_excerpt_maxlength
+             end
+    me = self.new(length, {})
     parser = Nokogiri::HTML::SAX::Parser.new(me)
     catch(:done) { parser.parse(html) }
     me.excerpt.strip

plugin.rb

@@ -16,7 +16,7 @@
     ../lib/discourse_activity_pub/request.rb
     ../lib/discourse_activity_pub/webfinger.rb
     ../lib/discourse_activity_pub/username_validator.rb
-    ../lib/discourse_activity_pub/excerpt_parser.rb
+    ../lib/discourse_activity_pub/content_parser.rb
     ../lib/discourse_activity_pub/signature_parser.rb
     ../lib/discourse_activity_pub/delivery_failure_tracker.rb
     ../lib/discourse_activity_pub/ap.rb
@@ -36,6 +36,7 @@
     ../lib/discourse_activity_pub/ap/activity/update.rb
     ../lib/discourse_activity_pub/ap/activity/undo.rb
     ../lib/discourse_activity_pub/ap/object/note.rb
+    ../lib/discourse_activity_pub/ap/object/article.rb
     ../lib/discourse_activity_pub/ap/collection.rb
     ../lib/discourse_activity_pub/ap/collection/ordered_collection.rb
     ../app/models/concerns/discourse_activity_pub/ap/identifier_validations.rb
@@ -72,6 +73,7 @@
     ../app/serializers/discourse_activity_pub/ap/actor/group_serializer.rb
     ../app/serializers/discourse_activity_pub/ap/actor/person_serializer.rb
     ../app/serializers/discourse_activity_pub/ap/object/note_serializer.rb
+    ../app/serializers/discourse_activity_pub/ap/object/article_serializer.rb
     ../app/serializers/discourse_activity_pub/ap/collection_serializer.rb
     ../app/serializers/discourse_activity_pub/ap/collection/ordered_collection_serializer.rb
     ../app/serializers/discourse_activity_pub/webfinger_serializer.rb
@@ -110,6 +112,7 @@
   register_category_custom_field_type("activity_pub_username", :string)
   register_category_custom_field_type("activity_pub_name", :string)
   register_category_custom_field_type("activity_pub_default_visibility", :string)
+  register_category_custom_field_type("activity_pub_post_object_type", :string)
   add_to_class(:category, :activity_pub_url) do
     "#{DiscourseActivityPub.base_url}#{self.url}"
   end
@@ -155,6 +158,12 @@
   add_to_class(:category, :activity_pub_default_visibility) do
     custom_fields["activity_pub_default_visibility"] || DiscourseActivityPubActivity::DEFAULT_VISIBILITY
   end
+  add_to_class(:category, :activity_pub_post_object_type) do
+    custom_fields["activity_pub_post_object_type"]
+  end
+  add_to_class(:category, :activity_pub_default_object_type) do
+    DiscourseActivityPub::AP::Actor::Group.type
+  end
 
   add_model_callback(:category, :after_save) do
     DiscourseActivityPubActor.ensure_for(self)
@@ -265,7 +274,7 @@
     if custom_fields["activity_pub_content"].present?
       custom_fields["activity_pub_content"]
     else
-      DiscourseActivityPub::ExcerptParser.get_content(self)
+      DiscourseActivityPub::ContentParser.get_content(self)
     end
   end
   add_to_class(:post, :activity_pub_actor) do
@@ -343,6 +352,13 @@
 
     performing_activity
   end
+  add_to_class(:post, :activity_pub_object_type) do
+    self.activity_pub_object&.ap_type || self.activity_pub_default_object_type
+  end
+  add_to_class(:post, :activity_pub_default_object_type) do
+    self.topic&.category&.activity_pub_post_object_type ||
+    DiscourseActivityPub::AP::Object::Note.type
+  end
 
   add_to_serializer(:post, :activity_pub_enabled) do
     object.activity_pub_enabled
@@ -359,7 +375,7 @@
     if post.activity_pub_enabled
       post.custom_fields[
         "activity_pub_content"
-      ] = DiscourseActivityPub::ExcerptParser.get_content(post)
+      ] = DiscourseActivityPub::ContentParser.get_content(post)
     end
   end
   on(:post_edited) do |post, topic_changed, post_revisor|
@@ -369,7 +385,7 @@
     if post.activity_pub_enabled
       post.custom_fields[
         "activity_pub_content"
-      ] = DiscourseActivityPub::ExcerptParser.get_content(post)
+      ] = DiscourseActivityPub::ContentParser.get_content(post)
       post.custom_fields[
         "activity_pub_visibility"
       ] = post_opts[:activity_pub_visibility] ||

spec/fabricators/discourse_activity_pub_object_fabricator.rb

@@ -16,3 +16,16 @@
     end
   end
 end
+
+Fabricator(:discourse_activity_pub_object_article, from: :discourse_activity_pub_object) do
+  ap_type { DiscourseActivityPub::AP::Object::Article.type }
+  model { Fabricate(:post) }
+  local { true }
+
+  after_create do |object|
+    if object.model.respond_to?(:activity_pub_content)
+      object.content = object.model.activity_pub_content
+      object.save!
+    end
+  end
+end

spec/lib/discourse_activity_pub/excerpt_parser_spec.rb → spec/lib/discourse_activity_pub/content_parser_spec.rb

@@ -1,17 +1,17 @@
 # frozen_string_literal: true
 
-RSpec.describe DiscourseActivityPub::ExcerptParser do
-  describe "#get_excerpt" do
+RSpec.describe DiscourseActivityPub::ContentParser do
+  describe "#get_note" do
     it "handles div note in short post" do
-      expect(described_class.get_excerpt("<div class='note'>hi</div> test", 100, {})).to eq("hi")
+      expect(described_class.get_note("<div class='note'>hi</div> test")).to eq("hi")
     end
 
     it "handles div note in long post" do
       html = <<~HTML
       <p>Lorem ipsum dolor sit amet, consectetur adipiscing elit. Duis quam nulla, feugiat venenatis elementum ut, imperdiet eu nisl. Vestibulum dictum luctus tortor, vel consequat lectus tristique non. Sed aliquam at eros et lacinia. Nam viverra libero at tortor semper fringilla non ut velit. Mauris dignissim sapien sed felis consequat, quis ullamcorper augue viverra. Donec elementum nisl ut leo viverra, vel consequat diam facilisis. Donec leo arcu, dictum vel vestibulum sit amet, maximus sed neque. Vestibulum blandit metus ante, sit amet porta lorem maximus id. Suspendisse sed lacus sapien. Nulla dui dui, dapibus vitae quam ut, elementum ultrices ipsum. In congue laoreet eleifend. Sed tincidunt consequat dolor, volutpat posuere arcu molestie a. Class aptent taciti sociosqu ad litora torquent per conubia nostra, per inceptos himenaeos.</p>
 <p>Pellentesque feugiat, elit ut aliquam fringilla, lectus eros rhoncus arcu, eget posuere ex velit ac purus. Morbi nec enim iaculis, lobortis lacus id, laoreet neque. Pellentesque et turpis a sapien tincidunt consequat quis a urna. Curabitur id ipsum vitae nisi dapibus tincidunt. Cras hendrerit nunc eget consectetur dapibus. Donec lacinia in sapien ac pellentesque. Phasellus at risus et lorem luctus pretium a eget leo. Nulla pellentesque metus libero, sit amet efficitur diam vehicula ac. Mauris ultrices erat non nulla volutpat tristique et sed arcu. </p><div class="note">hi</div><p></p>
       HTML
-      expect(described_class.get_excerpt(html, 100, {})).to eq("hi")
+      expect(described_class.get_note(html)).to eq("hi")
     end
   end
 
@@ -38,5 +38,13 @@
       post = Fabricate(:post, raw: content)
       expect(described_class.get_content(post)).to eq(content)
     end
+
+    context "with Article" do
+      it "returns all cooked content" do
+        Post.any_instance.stubs(:activity_pub_object_type).returns('Article')
+        post = Fabricate(:post_with_very_long_raw_content)
+        expect(described_class.get_content(post)).to eq(described_class.cook(post.raw))
+      end
+    end
   end
 end