about summary refs log tree commit diff
path: root/app
diff options
context:
space:
mode:
authorThibG <thib@sitedethib.com>2019-02-28 15:22:21 +0100
committerEugen Rochko <eugen@zeonfederated.com>2019-02-28 15:22:21 +0100
commit9d3c6f1849120e732a9230959cb302575765ea8c (patch)
tree9fea88d303fe39036ae87698bb723c44a0146811 /app
parent6e8743d17a92438bd2332e97f824bc212fa4b96e (diff)
Improved remote thread fetching (#10106)
* Fetch up to 5 replies when discovering a new remote status

This is used for resolving threads downwards. The originating
server must add a “replies” attributes with such replies for it to
be useful.

* Add some tests for ActivityPub::FetchRepliesWorker

* Add specs for ActivityPub::FetchRepliesService

* Serialize up to 5 public self-replies for ActivityPub notes

* Add specs for ActivityPub::NoteSerializer

* Move exponential backoff logic to a worker concern

* Fetch first page of paginated collections when fetching thread replies

* Add specs for paginated collections in replies

* Move Note replies serialization to a first CollectionPage

The collection isn't actually paginable yet as it has no id nor
a `next` field. This may come in another PR.

* Use pluck(:uri) instead of map(&:uri) to improve performances

* Fix fetching replies when they are in a CollectionPage
Diffstat (limited to 'app')
-rw-r--r--app/lib/activitypub/activity/create.rb10
-rw-r--r--app/models/concerns/status_threading_concern.rb4
-rw-r--r--app/presenters/activitypub/collection_presenter.rb2
-rw-r--r--app/serializers/activitypub/collection_serializer.rb5
-rw-r--r--app/serializers/activitypub/note_serializer.rb13
-rw-r--r--app/services/activitypub/fetch_replies_service.rb60
-rw-r--r--app/workers/activitypub/fetch_replies_worker.rb12
-rw-r--r--app/workers/concerns/exponential_backoff.rb11
-rw-r--r--app/workers/fetch_reply_worker.rb12
-rw-r--r--app/workers/thread_resolve_worker.rb5
10 files changed, 127 insertions, 7 deletions
diff --git a/app/lib/activitypub/activity/create.rb b/app/lib/activitypub/activity/create.rb
index 6d58aba70..0980f94ba 100644
--- a/app/lib/activitypub/activity/create.rb
+++ b/app/lib/activitypub/activity/create.rb
@@ -40,6 +40,7 @@ class ActivityPub::Activity::Create < ActivityPub::Activity
     end
 
     resolve_thread(@status)
+    fetch_replies(@status)
     distribute(@status)
     forward_for_reply if @status.public_visibility? || @status.unlisted_visibility?
   end
@@ -213,6 +214,15 @@ class ActivityPub::Activity::Create < ActivityPub::Activity
     ThreadResolveWorker.perform_async(status.id, in_reply_to_uri)
   end
 
+  def fetch_replies(status)
+    collection = @object['replies']
+    return if collection.nil?
+    replies = ActivityPub::FetchRepliesService.new.call(status, collection, false)
+    return if replies.present?
+    uri = value_or_id(collection)
+    ActivityPub::FetchRepliesWorker.perform_async(status.id, uri) unless uri.nil?
+  end
+
   def conversation_from_uri(uri)
     return nil if uri.nil?
     return Conversation.find_by(id: OStatus::TagManager.instance.unique_tag_to_local_id(uri, 'Conversation')) if OStatus::TagManager.instance.local_id?(uri)
diff --git a/app/models/concerns/status_threading_concern.rb b/app/models/concerns/status_threading_concern.rb
index b9c800c2a..15eb695cd 100644
--- a/app/models/concerns/status_threading_concern.rb
+++ b/app/models/concerns/status_threading_concern.rb
@@ -11,6 +11,10 @@ module StatusThreadingConcern
     find_statuses_from_tree_path(descendant_ids(limit, max_child_id, since_child_id, depth), account, promote: true)
   end
 
+  def self_replies(limit)
+    account.statuses.where(in_reply_to_id: id, visibility: [:public, :unlisted]).reorder(id: :asc).limit(limit)
+  end
+
   private
 
   def ancestor_ids(limit)
diff --git a/app/presenters/activitypub/collection_presenter.rb b/app/presenters/activitypub/collection_presenter.rb
index ec84ab1a3..28331f0c4 100644
--- a/app/presenters/activitypub/collection_presenter.rb
+++ b/app/presenters/activitypub/collection_presenter.rb
@@ -1,5 +1,5 @@
 # frozen_string_literal: true
 
 class ActivityPub::CollectionPresenter < ActiveModelSerializers::Model
-  attributes :id, :type, :size, :items, :part_of, :first, :last, :next, :prev
+  attributes :id, :type, :size, :items, :page, :part_of, :first, :last, :next, :prev
 end
diff --git a/app/serializers/activitypub/collection_serializer.rb b/app/serializers/activitypub/collection_serializer.rb
index e8960131b..b03609957 100644
--- a/app/serializers/activitypub/collection_serializer.rb
+++ b/app/serializers/activitypub/collection_serializer.rb
@@ -7,7 +7,8 @@ class ActivityPub::CollectionSerializer < ActiveModel::Serializer
     super
   end
 
-  attributes :id, :type
+  attribute :id, if: -> { object.id.present? }
+  attribute :type
   attribute :total_items, if: -> { object.size.present? }
   attribute :next, if: -> { object.next.present? }
   attribute :prev, if: -> { object.prev.present? }
@@ -37,6 +38,6 @@ class ActivityPub::CollectionSerializer < ActiveModel::Serializer
   end
 
   def page?
-    object.part_of.present?
+    object.part_of.present? || object.page.present?
   end
 end
diff --git a/app/serializers/activitypub/note_serializer.rb b/app/serializers/activitypub/note_serializer.rb
index c9d23e25f..6b0978ad3 100644
--- a/app/serializers/activitypub/note_serializer.rb
+++ b/app/serializers/activitypub/note_serializer.rb
@@ -13,6 +13,8 @@ class ActivityPub::NoteSerializer < ActiveModel::Serializer
   has_many :media_attachments, key: :attachment
   has_many :virtual_tags, key: :tag
 
+  has_one :replies, serializer: ActivityPub::CollectionSerializer
+
   def id
     ActivityPub::TagManager.instance.uri_for(object)
   end
@@ -33,6 +35,17 @@ class ActivityPub::NoteSerializer < ActiveModel::Serializer
     { object.language => Formatter.instance.format(object) }
   end
 
+  def replies
+    ActivityPub::CollectionPresenter.new(
+      type: :unordered,
+      first: ActivityPub::CollectionPresenter.new(
+        type: :unordered,
+        page: true,
+        items: object.self_replies(5).pluck(:uri)
+      )
+    )
+  end
+
   def language?
     object.language.present?
   end
diff --git a/app/services/activitypub/fetch_replies_service.rb b/app/services/activitypub/fetch_replies_service.rb
new file mode 100644
index 000000000..95c486a43
--- /dev/null
+++ b/app/services/activitypub/fetch_replies_service.rb
@@ -0,0 +1,60 @@
+# frozen_string_literal: true
+
+class ActivityPub::FetchRepliesService < BaseService
+  include JsonLdHelper
+
+  def call(parent_status, collection_or_uri, allow_synchronous_requests = true)
+    @account = parent_status.account
+    @allow_synchronous_requests = allow_synchronous_requests
+
+    @items = collection_items(collection_or_uri)
+    return if @items.nil?
+
+    FetchReplyWorker.push_bulk(filtered_replies)
+
+    @items
+  end
+
+  private
+
+  def collection_items(collection_or_uri)
+    collection = fetch_collection(collection_or_uri)
+    return unless collection.is_a?(Hash)
+
+    collection = fetch_collection(collection['first']) if collection['first'].present?
+    return unless collection.is_a?(Hash)
+
+    case collection['type']
+    when 'Collection', 'CollectionPage'
+      collection['items']
+    when 'OrderedCollection', 'OrderedCollectionPage'
+      collection['orderedItems']
+    end
+  end
+
+  def fetch_collection(collection_or_uri)
+    return collection_or_uri if collection_or_uri.is_a?(Hash)
+    return unless @allow_synchronous_requests
+    return if invalid_origin?(collection_or_uri)
+    collection = fetch_resource_without_id_validation(collection_or_uri)
+    raise Mastodon::UnexpectedResponseError if collection.nil?
+    collection
+  end
+
+  def filtered_replies
+    # Only fetch replies to the same server as the original status to avoid
+    # amplification attacks.
+
+    # Also limit to 5 fetched replies to limit potential for DoS.
+    @items.map { |item| value_or_id(item) }.reject { |uri| invalid_origin?(uri) }.take(5)
+  end
+
+  def invalid_origin?(url)
+    return true if unsupported_uri_scheme?(url)
+
+    needle   = Addressable::URI.parse(url).host
+    haystack = Addressable::URI.parse(@account.uri).host
+
+    !haystack.casecmp(needle).zero?
+  end
+end
diff --git a/app/workers/activitypub/fetch_replies_worker.rb b/app/workers/activitypub/fetch_replies_worker.rb
new file mode 100644
index 000000000..bf466db54
--- /dev/null
+++ b/app/workers/activitypub/fetch_replies_worker.rb
@@ -0,0 +1,12 @@
+# frozen_string_literal: true
+
+class ActivityPub::FetchRepliesWorker
+  include Sidekiq::Worker
+  include ExponentialBackoff
+
+  sidekiq_options queue: 'pull', retry: 3
+
+  def perform(parent_status_id, replies_uri)
+    ActivityPub::FetchRepliesService.new.call(Status.find(parent_status_id), replies_uri)
+  end
+end
diff --git a/app/workers/concerns/exponential_backoff.rb b/app/workers/concerns/exponential_backoff.rb
new file mode 100644
index 000000000..f2b931e33
--- /dev/null
+++ b/app/workers/concerns/exponential_backoff.rb
@@ -0,0 +1,11 @@
+# frozen_string_literal: true
+
+module ExponentialBackoff
+  extend ActiveSupport::Concern
+
+  included do
+    sidekiq_retry_in do |count|
+      15 + 10 * (count**4) + rand(10 * (count**4))
+    end
+  end
+end
diff --git a/app/workers/fetch_reply_worker.rb b/app/workers/fetch_reply_worker.rb
new file mode 100644
index 000000000..f7aa25e81
--- /dev/null
+++ b/app/workers/fetch_reply_worker.rb
@@ -0,0 +1,12 @@
+# frozen_string_literal: true
+
+class FetchReplyWorker
+  include Sidekiq::Worker
+  include ExponentialBackoff
+
+  sidekiq_options queue: 'pull', retry: 3
+
+  def perform(child_url)
+    FetchRemoteStatusService.new.call(child_url)
+  end
+end
diff --git a/app/workers/thread_resolve_worker.rb b/app/workers/thread_resolve_worker.rb
index c18a778d5..8bba9ca75 100644
--- a/app/workers/thread_resolve_worker.rb
+++ b/app/workers/thread_resolve_worker.rb
@@ -2,13 +2,10 @@
 
 class ThreadResolveWorker
   include Sidekiq::Worker
+  include ExponentialBackoff
 
   sidekiq_options queue: 'pull', retry: 3
 
-  sidekiq_retry_in do |count|
-    15 + 10 * (count**4) + rand(10 * (count**4))
-  end
-
   def perform(child_status_id, parent_url)
     child_status  = Status.find(child_status_id)
     parent_status = FetchRemoteStatusService.new.call(parent_url)