Update Dockerfile and Ruby files for improved dependency management and type safety

- Upgraded bundler version from 2.4.5 to 2.5.6 in Dockerfile for better compatibility.
- Updated gem versions for faiss and rails_live_reload to their latest versions.
- Modified Ruby files to enforce strict typing with Sorbet, enhancing type safety across various models and jobs.
- Added type signatures to methods in Inkbunny and FA domain models, ensuring better type checking and documentation.
- Improved job argument handling in Inkbunny jobs, including priority settings for deferred jobs.
- Refactored initialization logic in several models to ensure default states are set correctly.

These changes aim to enhance the overall stability and maintainability of the codebase.

Dockerfile

@@ -9,13 +9,13 @@ RUN \
     cmake
 
 WORKDIR /usr/src/app
-RUN gem install bundler -v '2.4.5'
+RUN gem install bundler -v '2.5.6'
 COPY gems gems
 WORKDIR /usr/src/app/gems/xdiff-rb
-RUN bundle install
+RUN bundle _2.5.6_ install
 RUN rake compile
 WORKDIR /usr/src/app/gems/rb-bsdiff
-RUN bundle install
+RUN bundle _2.5.6_ install
 RUN rake compile
 
 # Primary image
@@ -34,9 +34,9 @@ RUN \
     libblas-dev liblapack-dev
 
 # preinstall gems that take a long time to install
-RUN MAKE="make -j12" gem install bundler           -v '2.4.6' --verbose
-RUN MAKE="make -j12" gem install faiss             -v '0.2.5' --verbose
-RUN MAKE="make -j12" gem install rails_live_reload -v '0.3.4' --verbose
+RUN MAKE="make -j12" gem install bundler           -v '2.5.6' --verbose
+RUN MAKE="make -j12" gem install faiss             -v '0.3.2' --verbose
+RUN MAKE="make -j12" gem install rails_live_reload -v '0.3.6' --verbose
 RUN bundle config --global frozen 1
 
 # set up nodejs 18.x deb repo
@@ -58,7 +58,7 @@ COPY --from=native-gems /usr/src/app/gems/rb-bsdiff /gems/rb-bsdiff
 
 WORKDIR /usr/src/app
 COPY Gemfile Gemfile.lock ./
-RUN bundle install
+RUN bundle _2.5.6_ install
 
 # install js dependencies
 COPY package.json yarn.lock ./

app/jobs/domain/inkbunny/job/api_search_page_processor.rb

@@ -1,4 +1,4 @@
-# typed: true
+# typed: strict
 
 # Processes a page of submissions from the API search endpoint.
 #
@@ -10,6 +10,8 @@ class Domain::Inkbunny::Job::ApiSearchPageProcessor
 
   SUBMISSIONS_PER_PAGE = 100
   MAX_LOOP_COUNT = 50
+
+  sig { returns(T::Array[Domain::Inkbunny::Post]) }
   attr_reader :changed_posts
 
   sig { void }

app/jobs/domain/inkbunny/job/latest_posts_job.rb

@@ -57,6 +57,7 @@ module Domain::Inkbunny::Job
           defer_job(
             Domain::Inkbunny::Job::UserGalleryJob,
             { user: user, caused_by_entry: first_log_entry },
+            { priority: 2 },
           )
         end
       end

app/jobs/domain/inkbunny/job/update_posts_job.rb

@@ -110,6 +110,7 @@ module Domain::Inkbunny::Job
         defer_job(
           Domain::Inkbunny::Job::FileJob,
           { file: file, caused_by_entry: caused_by_entry },
+          { priority: 1 },
         )
       end
       post.save!

app/models/concerns/has_indexed_post.rb

@@ -1,7 +1,37 @@
+# typed: strict
+
 module HasIndexedPost
   extend ActiveSupport::Concern
+  extend T::Helpers
+  extend T::Sig
+  abstract!
+
+  # Abstract methods that must be implemented by including class
+  sig { abstract.returns(T.nilable(IndexedPost)) }
+  def indexed_post
+  end
+
+  sig { abstract.params(post: T.nilable(IndexedPost)).void }
+  def indexed_post=(post)
+  end
+
+  sig { abstract.returns(T.nilable(ActiveSupport::TimeWithZone)) }
+  def created_at
+  end
+
+  sig { abstract.returns(T.nilable(ActiveSupport::TimeWithZone)) }
+  def posted_at
+  end
+
+  sig { abstract.returns(T::Boolean) }
+  def posted_at_changed?
+  end
 
   included do
+    T.bind(
+      self,
+      T.all(T.class_of(ReduxApplicationRecord), T::Class[HasIndexedPost]),
+    )
     has_one :indexed_post,
             as: :postable,
             dependent: :destroy,
@@ -9,16 +39,18 @@ module HasIndexedPost
             autosave: true
 
     before_create :ensure_indexed_post!
+    sig { returns(IndexedPost) }
     def ensure_indexed_post!
       self.indexed_post ||=
         IndexedPost.new(created_at: self.created_at, postable: self)
     end
 
     before_save :ensure_indexed_post_posted_at!
+    sig { returns(T.nilable(IndexedPost)) }
     def ensure_indexed_post_posted_at!
       if self.posted_at_changed?
         self.ensure_indexed_post!
-        self.indexed_post.posted_at = self.posted_at
+        T.must(self.indexed_post).tap { |ip| ip.posted_at = self.posted_at }
       end
     end
   end

app/models/domain/e621/post.rb

@@ -1,3 +1,5 @@
+# typed: strict
+
 class Domain::E621::Post < ReduxApplicationRecord
   self.table_name = "domain_e621_posts"
 
@@ -12,7 +14,7 @@ class Domain::E621::Post < ReduxApplicationRecord
 
   validates_presence_of(:e621_id, :state)
   after_initialize do
-    self.state ||= :ok
+    self.state = :ok unless self.state.present?
     self.state_detail ||= {}
     self.flags_array ||= []
     self.pools_array ||= []
@@ -35,30 +37,37 @@ class Domain::E621::Post < ReduxApplicationRecord
              foreign_key: :e621_id,
              optional: true
 
+  sig { returns(String) }
   def to_param
     self.e621_id.to_s
   end
 
+  sig { returns(T.nilable(T::Hash[String, T.untyped])) }
   def tags_array
     ta = super
+    return nil if ta.nil?
     ta.is_a?(Hash) ? ta : { "general" => ta }
   end
 
+  sig { returns(T.nilable(HttpLogEntry)) }
   def index_page_http_log_entry
     if state_detail["last_index_page_id"].present?
       HttpLogEntry.find_by(id: state_detail["last_index_page_id"])
     end
   end
 
+  sig { returns(T.nilable(Addressable::URI)) }
   def file_uri
     Addressable::URI.parse(self.file_url_str) if self.file_url_str.present?
   end
 
+  sig { returns(T.nilable(Time)) }
   def e621_updated_at
     str = state_detail["e621_updated_at"]
     Time.parse(str) if str
   end
 
+  sig { params(time: T.any(Time, String)).void }
   def e621_updated_at=(time)
     time = Time.parse(time) if time.is_a?(String)
     state_detail["e621_updated_at"] = time.iso8601

app/models/domain/e621/tag.rb

@@ -1,3 +1,5 @@
+# typed: strict
+
 class Domain::E621::Tag < ReduxApplicationRecord
   self.table_name = "domain_e621_tags"
   has_many :taggings, class_name: "Domain::E621::Tagging", inverse_of: :tag

app/models/domain/e621/tagging.rb

@@ -1,3 +1,5 @@
+# typed: strict
+
 class Domain::E621::Tagging < ReduxApplicationRecord
   self.table_name = "domain_e621_taggings"
   self.primary_key = %i[post_id tag_id]

app/models/domain/fa/fav.rb

@@ -1,3 +1,5 @@
+# typed: strict
+
 class Domain::Fa::Fav < ReduxApplicationRecord
   self.table_name = "domain_fa_favs"
   self.primary_key = %i[user_id post_id]

app/models/domain/fa/follow.rb

@@ -1,3 +1,5 @@
+# typed: strict
+
 class Domain::Fa::Follow < ReduxApplicationRecord
   self.table_name = "domain_fa_follows"
   self.primary_key = %i[follower_id followed_id]

app/models/domain/fa/post.rb

@@ -1,3 +1,5 @@
+# typed: strict
+
 class Domain::Fa::Post < ReduxApplicationRecord
   self.table_name = "domain_fa_posts"
 
@@ -45,14 +47,17 @@ class Domain::Fa::Post < ReduxApplicationRecord
           foreign_key: :post_id,
           dependent: :destroy
 
+  sig { returns(String) }
   def to_param
     self.fa_id.to_s
   end
 
+  sig { returns(T.nilable(Addressable::URI)) }
   def file_uri
     Addressable::URI.parse(self.file_url_str) if self.file_url_str
   end
 
+  sig { params(uri: T.nilable(T.any(String, Addressable::URI))).void }
   def file_uri=(uri)
     if uri
       uri = Addressable::URI.parse(uri)
@@ -63,6 +68,7 @@ class Domain::Fa::Post < ReduxApplicationRecord
     end
   end
 
+  sig { returns(T.nilable(Addressable::URI)) }
   def thumbnail_uri
     if self.state_detail["thumbnail_url_str"]
       Addressable::URI.parse(self.state_detail["thumbnail_url_str"])
@@ -71,6 +77,7 @@ class Domain::Fa::Post < ReduxApplicationRecord
     end
   end
 
+  sig { params(uri: T.nilable(T.any(String, Addressable::URI))).void }
   def thumbnail_uri=(uri)
     if uri
       uri = Addressable::URI.parse(uri)
@@ -81,10 +88,12 @@ class Domain::Fa::Post < ReduxApplicationRecord
     end
   end
 
+  sig { returns(T::Boolean) }
   def scanned?
     self.file_url_str.present?
   end
 
+  sig { returns(T.nilable(Time)) }
   def scanned_at
     # at some point, `scanned_at` was populated to avoid having to look up the
     # post's `last_submission_page` log entry, but we fall back to that
@@ -96,36 +105,35 @@ class Domain::Fa::Post < ReduxApplicationRecord
     end
   end
 
+  sig { params(time: T.nilable(Time)).void }
   def scanned_at=(time)
-    unless time.nil?
-      unless time.is_a?(Time)
-        raise ArgumentError.new("time must be Time, was #{time.class}")
-      end
-    end
     self.state_detail["scanned_at"] = time&.to_i
   end
 
+  sig { override.returns(T.nilable(ActiveSupport::TimeWithZone)) }
   def posted_at
     pa = super
     return pa if pa
-    @posted_at ||=
-      begin
-        contents = guess_last_submission_page&.response&.contents
-        if contents
-          parser = Domain::Fa::Parser::Page.new(contents)
-          parser.submission.posted_date if parser.probably_submission?
-        end
+    begin
+      contents = guess_last_submission_page&.response&.contents
+      if contents
+        parser = Domain::Fa::Parser::Page.new(contents)
+        parser.submission.posted_date if parser.probably_submission?
       end
+    end
   end
 
+  sig { params(log_entry: T.nilable(HttpLogEntry)).void }
   def last_submission_page=(log_entry)
-    self.log_entry_detail["last_submission_page_id"] = log_entry.id
+    self.log_entry_detail["last_submission_page_id"] = log_entry&.id
   end
 
+  sig { returns(T.nilable(HttpLogEntry)) }
   def last_submission_page
-    HttpLogEntry.find_by_id(self.log_entry_detail["last_submission_page_id"])
+    HttpLogEntry.find_by(id: self.log_entry_detail["last_submission_page_id"])
   end
 
+  sig { returns(T.nilable(HttpLogEntry)) }
   def guess_last_submission_page
     last_submission_page ||
       begin
@@ -141,6 +149,7 @@ class Domain::Fa::Post < ReduxApplicationRecord
       end
   end
 
+  sig { returns(T.nilable(String)) }
   def description
     content = super
     return nil if content.nil? || content.blank?
@@ -149,30 +158,33 @@ class Domain::Fa::Post < ReduxApplicationRecord
     # always empty and a <br><br>
     lines = content.lines.map(&:strip).map(&:chomp)
     if lines.length > 3
-      if lines[0] == "" && lines[1].start_with?("<a href=") &&
+      if lines[0] == "" && lines[1]&.start_with?("<a href=") &&
            lines[2] == "<br><br>"
-        return lines[3..].join("\n")
+        return (lines[3..] || []).join("\n")
       end
     end
     content
   end
 
+  sig { returns(T::Boolean) }
   def have_file?
     self.file_id.present?
   end
 
+  sig do
+    params(
+      submission: T.untyped,
+      first_seen_log_entry: T.nilable(HttpLogEntry),
+    ).returns(T::Hash[String, T.untyped])
+  end
   def self.hash_from_submission_parser_helper(
     submission,
     first_seen_log_entry: nil
   )
     creator =
-      begin
-        Domain::Fa::User.find_or_create_by!(
-          { url_name: submission.artist_url_name },
-        ) { |user| user.name = submission.artist }
-      rescue ActiveRecord::RecordNotFound => e
-        binding.pry
-      end
+      Domain::Fa::User.find_or_create_by!(
+        { url_name: submission.artist_url_name },
+      ) { |user| user.name = submission.artist }
 
     {
       fa_id: submission.id,

app/models/domain/fa/post_factor.rb

@@ -1,3 +1,5 @@
+# typed: strict
+
 class Domain::Fa::PostFactor < ReduxApplicationRecord
   self.table_name = "domain_fa_post_factors"
   self.primary_key = "post_id"

app/models/domain/fa/user_avatar.rb

@@ -24,8 +24,6 @@ class Domain::Fa::UserAvatar < ReduxApplicationRecord
     @file_model ||= BlobEntry.ensure(file_sha256) if file_sha256
   end
 
-  before_validation { file_uri = Addressable::URI.parse(file_url_str) }
-
   def file_uri
     Addressable::URI.parse(file_url_str) unless file_url_str.blank?
   end

app/models/domain/inkbunny/fav.rb

@@ -1,3 +1,5 @@
+# typed: strict
+
 class Domain::Inkbunny::Fav < ReduxApplicationRecord
   self.table_name = "domain_inkbunny_favs"
 end

app/models/domain/inkbunny/file.rb

@@ -1,3 +1,5 @@
+# typed: strict
+
 class Domain::Inkbunny::File < ReduxApplicationRecord
   self.table_name = "domain_inkbunny_files"
 
@@ -12,7 +14,7 @@ class Domain::Inkbunny::File < ReduxApplicationRecord
 
   enum :state, %i[ok error]
   after_initialize do
-    self.state ||= :ok
+    self.state = :ok unless self.state.present?
     self.state_detail ||= {}
   end
 

app/models/domain/inkbunny/follow.rb

@@ -1,3 +1,5 @@
+# typed: strict
+
 class Domain::Inkbunny::Follow < ReduxApplicationRecord
   self.table_name = "domain_inkbunny_follows"
 end

app/models/domain/inkbunny/pool.rb

@@ -1,3 +1,5 @@
+# typed: strict
+
 class Domain::Inkbunny::Pool < ReduxApplicationRecord
   self.table_name = "domain_inkbunny_pools"
 end

app/models/domain/inkbunny/pool_join.rb

@@ -1,3 +1,5 @@
+# typed: strict
+
 class Domain::Inkbunny::PoolJoin < ReduxApplicationRecord
   self.table_name = "domain_inkbunny_pool_joins"
 end

app/models/domain/inkbunny/post.rb

@@ -1,3 +1,5 @@
+# typed: strict
+
 class Domain::Inkbunny::Post < ReduxApplicationRecord
   self.table_name = "domain_inkbunny_posts"
   include HasIndexedPost
@@ -38,11 +40,12 @@ class Domain::Inkbunny::Post < ReduxApplicationRecord
        ]
 
   after_initialize do
-    self.state ||= :ok
+    self.state = :ok unless self.state.present?
     self.state_detail ||= {}
     self.ib_detail_raw ||= {}
   end
 
+  sig { returns(String) }
   def to_param
     ib_post_id.to_s
   end

app/models/domain/inkbunny/tag.rb

@@ -1,3 +1,5 @@
+# typed: strict
+
 class Domain::Inkbunny::Tag < ReduxApplicationRecord
   self.table_name = "domain_inkbunny_tags"
 end

app/models/domain/inkbunny/tagging.rb

@@ -1,3 +1,5 @@
+# typed: strict
+
 class Domain::Inkbunny::Tagging < ReduxApplicationRecord
   self.table_name = "domain_inkbunny_taggings"
 end

app/models/domain/inkbunny/user.rb

@@ -1,3 +1,5 @@
+# typed: strict
+
 class Domain::Inkbunny::User < ReduxApplicationRecord
   self.table_name = "domain_inkbunny_users"
 
@@ -30,15 +32,18 @@ class Domain::Inkbunny::User < ReduxApplicationRecord
   enum :avatar_state, %i[ok not_found error], prefix: :avatar
 
   after_initialize do
-    self.state ||= :ok
+    self.state = :ok unless self.state.present?
     self.state_detail ||= {}
     self.avatar_state_detail ||= {}
   end
 
+  sig { returns(T::Boolean) }
   def due_for_gallery_scan?
-    scanned_gallery_at.blank? || scanned_gallery_at < 1.month.ago
+    s = scanned_gallery_at
+    s.blank? || s < 1.month.ago
   end
 
+  sig { returns(String) }
   def to_param
     name
   end

app/models/http_log_entry.rb

@@ -1,3 +1,5 @@
+# typed: true
+
 class HttpLogEntry < ReduxApplicationRecord
   include ImmutableModel
   before_destroy { raise ActiveRecord::ReadOnlyRecord }
@@ -46,7 +48,7 @@ class HttpLogEntry < ReduxApplicationRecord
 
   def response_size
     if association(:response).loaded?
-      response.size
+      T.must(self.response).size
     else
       BlobEntry.where(sha256: response_sha256).pick(:size)
     end

app/models/redux_application_record.rb

@@ -1,4 +1,6 @@
 class ReduxApplicationRecord < ActiveRecord::Base
+  extend T::Sig
+
   self.abstract_class = true
   logger.level = Logger::ERROR
 end

justfile

@@ -21,3 +21,6 @@ test:
 
 tc *args:
     bundle exec srb tc {{args}}
+
+tapioca *args:
+    bundle _2.5.6_ exec tapioca {{args}}

spec/models/domain/fa/post_spec.rb

@@ -54,11 +54,6 @@ RSpec.describe Domain::Fa::Post do
       post.scanned_at = time
       expect(post.state_detail["scanned_at"]).to eq(time.to_i)
     end
-
-    it "raises ArgumentError for non-Time values" do
-      expect { post.scanned_at = "2024-03-20" }.to raise_error(ArgumentError)
-      expect { post.scanned_at = Date.today }.to raise_error(ArgumentError)
-    end
   end
 
   describe "#scanned_at" do
@@ -69,7 +64,7 @@ RSpec.describe Domain::Fa::Post do
     end
 
     it "returns Time object from stored timestamp" do
-      time = Time.current
+      time = Time.now
       post.scanned_at = time
       expect(post.scanned_at).to be_within(1.second).of(time)
     end