require 'carrierwave/orm/activerecord'

class Project < ActiveRecord::Base
  include Gitlab::ConfigHelper
  include Gitlab::ShellAdapter
  include Gitlab::VisibilityLevel
  include Gitlab::CurrentSettings
  include AccessRequestable
  include Avatarable
  include CacheMarkdownField
  include Referable
  include Sortable
  include AfterCommitQueue
  include CaseSensitivity
  include TokenAuthenticatable
  include ValidAttribute
  include ProjectFeaturesCompatibility
  include SelectForProjectAuthorization
  include Routable

  extend Gitlab::ConfigHelper
  extend Gitlab::CurrentSettings

  BoardLimitExceeded = Class.new(StandardError)

  NUMBER_OF_PERMITTED_BOARDS = 1
  UNKNOWN_IMPORT_URL = 'http://unknown.git'.freeze
  LATEST_STORAGE_VERSION = 1

  cache_markdown_field :description, pipeline: :description

  delegate :feature_available?, :builds_enabled?, :wiki_enabled?,
           :merge_requests_enabled?, :issues_enabled?, to: :project_feature,
                                                       allow_nil: true

  delegate :base_dir, :disk_path, :ensure_storage_path_exists, to: :storage

  default_value_for :archived, false
  default_value_for :visibility_level, gitlab_config_features.visibility_level
  default_value_for :resolve_outdated_diff_discussions, false
  default_value_for :container_registry_enabled, gitlab_config_features.container_registry
  default_value_for(:repository_storage) { current_application_settings.pick_repository_storage }
  default_value_for(:shared_runners_enabled) { current_application_settings.shared_runners_enabled }
  default_value_for :issues_enabled, gitlab_config_features.issues
  default_value_for :merge_requests_enabled, gitlab_config_features.merge_requests
  default_value_for :builds_enabled, gitlab_config_features.builds
  default_value_for :wiki_enabled, gitlab_config_features.wiki
  default_value_for :snippets_enabled, gitlab_config_features.snippets
  default_value_for :only_allow_merge_if_all_discussions_are_resolved, false

  add_authentication_token_field :runners_token
  before_save :ensure_runners_token

  after_save :update_project_statistics, if: :namespace_id_changed?
  after_create :create_project_feature, unless: :project_feature
  after_create :set_last_activity_at
  after_create :set_last_repository_updated_at
  after_update :update_forks_visibility_level

  before_destroy :remove_private_deploy_keys
  after_destroy -> { run_after_commit { remove_pages } }

  after_validation :check_pending_delete

  # Storage specific hooks
  after_initialize :use_hashed_storage
  after_create :check_repository_absence!
  after_create :ensure_storage_path_exists
  after_save :ensure_storage_path_exists, if: :namespace_id_changed?

  acts_as_taggable

  attr_accessor :old_path_with_namespace
  attr_accessor :template_name
  attr_writer :pipeline_status
  attr_accessor :skip_disk_validation

  alias_attribute :title, :name

  # Relations
  belongs_to :creator, class_name: 'User'
  belongs_to :group, -> { where(type: 'Group') }, foreign_key: 'namespace_id'
  belongs_to :namespace

  has_one :last_event, -> {order 'events.created_at DESC'}, class_name: 'Event'
  has_many :boards, before_add: :validate_board_limit

  # Project services
  has_one :campfire_service
  has_one :drone_ci_service
  has_one :emails_on_push_service
  has_one :pipelines_email_service
  has_one :irker_service
  has_one :pivotaltracker_service
  has_one :hipchat_service
  has_one :flowdock_service
  has_one :assembla_service
  has_one :asana_service
  has_one :gemnasium_service
  has_one :mattermost_slash_commands_service
  has_one :mattermost_service
  has_one :slack_slash_commands_service
  has_one :slack_service
  has_one :buildkite_service
  has_one :bamboo_service
  has_one :teamcity_service
  has_one :pushover_service
  has_one :jira_service
  has_one :redmine_service
  has_one :custom_issue_tracker_service
  has_one :bugzilla_service
  has_one :gitlab_issue_tracker_service, inverse_of: :project
  has_one :external_wiki_service
  has_one :kubernetes_service, inverse_of: :project
  has_one :prometheus_service, inverse_of: :project
  has_one :mock_ci_service
  has_one :mock_deployment_service
  has_one :mock_monitoring_service
  has_one :microsoft_teams_service

  # TODO: replace these relations with the fork network versions
  has_one  :forked_project_link,  foreign_key: "forked_to_project_id"
  has_one  :forked_from_project,  through:   :forked_project_link

  has_many :forked_project_links, foreign_key: "forked_from_project_id"
  has_many :forks,                through:     :forked_project_links, source: :forked_to_project
  # TODO: replace these relations with the fork network versions

  has_one :root_of_fork_network,
          foreign_key: 'root_project_id',
          inverse_of: :root_project,
          class_name: 'ForkNetwork'
  has_one :fork_network_member
  has_one :fork_network, through: :fork_network_member

  # Merge Requests for target project should be removed with it
  has_many :merge_requests, foreign_key: 'target_project_id'
  has_many :issues
  has_many :labels, class_name: 'ProjectLabel'
  has_many :services
  has_many :events
  has_many :milestones
  has_many :notes
  has_many :snippets, class_name: 'ProjectSnippet'
  has_many :hooks, class_name: 'ProjectHook'
  has_many :protected_branches
  has_many :protected_tags

  has_many :project_authorizations
  has_many :authorized_users, through: :project_authorizations, source: :user, class_name: 'User'
  has_many :project_members, -> { where(requested_at: nil) },
    as: :source, dependent: :delete_all # rubocop:disable Cop/ActiveRecordDependent

  alias_method :members, :project_members
  has_many :users, through: :project_members

  has_many :requesters, -> { where.not(requested_at: nil) },
    as: :source, class_name: 'ProjectMember', dependent: :delete_all # rubocop:disable Cop/ActiveRecordDependent
  has_many :members_and_requesters, as: :source, class_name: 'ProjectMember'

  has_many :deploy_keys_projects
  has_many :deploy_keys, through: :deploy_keys_projects
  has_many :users_star_projects
  has_many :starrers, through: :users_star_projects, source: :user
  has_many :releases
  has_many :lfs_objects_projects, dependent: :destroy # rubocop:disable Cop/ActiveRecordDependent
  has_many :lfs_objects, through: :lfs_objects_projects
  has_many :project_group_links
  has_many :invited_groups, through: :project_group_links, source: :group
  has_many :pages_domains
  has_many :todos
  has_many :notification_settings, as: :source, dependent: :delete_all # rubocop:disable Cop/ActiveRecordDependent

  has_one :import_data, class_name: 'ProjectImportData', inverse_of: :project, autosave: true
  has_one :project_feature, inverse_of: :project
  has_one :statistics, class_name: 'ProjectStatistics'
  has_one :cluster, class_name: 'Gcp::Cluster', inverse_of: :project

  # Container repositories need to remove data from the container registry,
  # which is not managed by the DB. Hence we're still using dependent: :destroy
  # here.
  has_many :container_repositories, dependent: :destroy # rubocop:disable Cop/ActiveRecordDependent

  has_many :commit_statuses
  has_many :pipelines, class_name: 'Ci::Pipeline'

  # Ci::Build objects store data on the file system such as artifact files and
  # build traces. Currently there's no efficient way of removing this data in
  # bulk that doesn't involve loading the rows into memory. As a result we're
  # still using `dependent: :destroy` here.
  has_many :builds, class_name: 'Ci::Build', dependent: :destroy # rubocop:disable Cop/ActiveRecordDependent
  has_many :build_trace_section_names, class_name: 'Ci::BuildTraceSectionName'
  has_many :runner_projects, class_name: 'Ci::RunnerProject'
  has_many :runners, through: :runner_projects, source: :runner, class_name: 'Ci::Runner'
  has_many :variables, class_name: 'Ci::Variable'
  has_many :triggers, class_name: 'Ci::Trigger'
  has_many :environments
  has_many :deployments
  has_many :pipeline_schedules, class_name: 'Ci::PipelineSchedule'

  has_many :active_runners, -> { active }, through: :runner_projects, source: :runner, class_name: 'Ci::Runner'

  has_one :auto_devops, class_name: 'ProjectAutoDevops'

  accepts_nested_attributes_for :variables, allow_destroy: true
  accepts_nested_attributes_for :project_feature, update_only: true
  accepts_nested_attributes_for :import_data
  accepts_nested_attributes_for :auto_devops, update_only: true

  delegate :name, to: :owner, allow_nil: true, prefix: true
  delegate :members, to: :team, prefix: true
  delegate :add_user, :add_users, to: :team
  delegate :add_guest, :add_reporter, :add_developer, :add_master, to: :team
  delegate :empty_repo?, to: :repository

  # Validations
  validates :creator, presence: true, on: :create
  validates :description, length: { maximum: 2000 }, allow_blank: true
  validates :ci_config_path,
    format: { without: /\.{2}/,
              message: 'cannot include directory traversal.' },
    length: { maximum: 255 },
    allow_blank: true
  validates :name,
    presence: true,
    length: { maximum: 255 },
    format: { with: Gitlab::Regex.project_name_regex,
              message: Gitlab::Regex.project_name_regex_message }
  validates :path,
    presence: true,
    dynamic_path: true,
    length: { maximum: 255 },
    format: { with: Gitlab::PathRegex.project_path_format_regex,
              message: Gitlab::PathRegex.project_path_format_message },
    uniqueness: { scope: :namespace_id }

  validates :namespace, presence: true
  validates :name, uniqueness: { scope: :namespace_id }
  validates :import_url, addressable_url: true, if: :external_import?
  validates :import_url, importable_url: true, if: [:external_import?, :import_url_changed?]
  validates :star_count, numericality: { greater_than_or_equal_to: 0 }
  validate :check_limit, on: :create
  validate :check_repository_path_availability, on: :update, if: ->(project) { project.renamed? }
  validate :avatar_type,
    if: ->(project) { project.avatar.present? && project.avatar_changed? }
  validates :avatar, file_size: { maximum: 200.kilobytes.to_i }
  validate :visibility_level_allowed_by_group
  validate :visibility_level_allowed_as_fork
  validate :check_wiki_path_conflict
  validates :repository_storage,
    presence: true,
    inclusion: { in: ->(_object) { Gitlab.config.repositories.storages.keys } }

  mount_uploader :avatar, AvatarUploader
  has_many :uploads, as: :model, dependent: :destroy # rubocop:disable Cop/ActiveRecordDependent

  # Scopes
  scope :pending_delete, -> { where(pending_delete: true) }
  scope :without_deleted, -> { where(pending_delete: false) }

  scope :with_hashed_storage, -> { where('storage_version >= 1') }
  scope :with_legacy_storage, -> { where(storage_version: [nil, 0]) }

  scope :sorted_by_activity, -> { reorder(last_activity_at: :desc) }
  scope :sorted_by_stars, -> { reorder('projects.star_count DESC') }

  scope :in_namespace, ->(namespace_ids) { where(namespace_id: namespace_ids) }
  scope :personal, ->(user) { where(namespace_id: user.namespace_id) }
  scope :joined, ->(user) { where('namespace_id != ?', user.namespace_id) }
  scope :starred_by, ->(user) { joins(:users_star_projects).where('users_star_projects.user_id': user.id) }
  scope :visible_to_user, ->(user) { where(id: user.authorized_projects.select(:id).reorder(nil)) }
  scope :archived, -> { where(archived: true) }
  scope :non_archived, -> { where(archived: false) }
  scope :for_milestones, ->(ids) { joins(:milestones).where('milestones.id' => ids).distinct }
  scope :with_push, -> { joins(:events).where('events.action = ?', Event::PUSHED) }

  scope :with_project_feature, -> { joins('LEFT JOIN project_features ON projects.id = project_features.project_id') }
  scope :with_statistics, -> { includes(:statistics) }
  scope :with_shared_runners, -> { where(shared_runners_enabled: true) }
  scope :inside_path, ->(path) do
    # We need routes alias rs for JOIN so it does not conflict with
    # includes(:route) which we use in ProjectsFinder.
    joins("INNER JOIN routes rs ON rs.source_id = projects.id AND rs.source_type = 'Project'")
      .where('rs.path LIKE ?', "#{sanitize_sql_like(path)}/%")
  end

  # "enabled" here means "not disabled". It includes private features!
  scope :with_feature_enabled, ->(feature) {
    access_level_attribute = ProjectFeature.access_level_attribute(feature)
    with_project_feature.where(project_features: { access_level_attribute => [nil, ProjectFeature::PRIVATE, ProjectFeature::ENABLED] })
  }

  # Picks a feature where the level is exactly that given.
  scope :with_feature_access_level, ->(feature, level) {
    access_level_attribute = ProjectFeature.access_level_attribute(feature)
    with_project_feature.where(project_features: { access_level_attribute => level })
  }

  scope :with_builds_enabled, -> { with_feature_enabled(:builds) }
  scope :with_issues_enabled, -> { with_feature_enabled(:issues) }
  scope :with_merge_requests_enabled, -> { with_feature_enabled(:merge_requests) }

  enum auto_cancel_pending_pipelines: { disabled: 0, enabled: 1 }

  # Returns a collection of projects that is either public or visible to the
  # logged in user.
  def self.public_or_visible_to_user(user = nil)
    if user
      authorized = user
        .project_authorizations
        .select(1)
        .where('project_authorizations.project_id = projects.id')

      levels = Gitlab::VisibilityLevel.levels_for_user(user)

      where('EXISTS (?) OR projects.visibility_level IN (?)', authorized, levels)
    else
      public_to_user
    end
  end

  # project features may be "disabled", "internal" or "enabled". If "internal",
  # they are only available to team members. This scope returns projects where
  # the feature is either enabled, or internal with permission for the user.
  #
  # This method uses an optimised version of `with_feature_access_level` for
  # logged in users to more efficiently get private projects with the given
  # feature.
  def self.with_feature_available_for_user(feature, user)
    visible = [nil, ProjectFeature::ENABLED]

    if user&.admin?
      with_feature_enabled(feature)
    elsif user
      column = ProjectFeature.quoted_access_level_column(feature)

      authorized = user.project_authorizations.select(1)
        .where('project_authorizations.project_id = projects.id')

      with_project_feature
        .where("#{column} IN (?) OR (#{column} = ? AND EXISTS (?))",
              visible,
              ProjectFeature::PRIVATE,
              authorized)
    else
      with_feature_access_level(feature, visible)
    end
  end

  scope :active, -> { joins(:issues, :notes, :merge_requests).order('issues.created_at, notes.created_at, merge_requests.created_at DESC') }
  scope :abandoned, -> { where('projects.last_activity_at < ?', 6.months.ago) }

  scope :excluding_project, ->(project) { where.not(id: project) }

  state_machine :import_status, initial: :none do
    event :import_schedule do
      transition [:none, :finished, :failed] => :scheduled
    end

    event :force_import_start do
      transition [:none, :finished, :failed] => :started
    end

    event :import_start do
      transition scheduled: :started
    end

    event :import_finish do
      transition started: :finished
    end

    event :import_fail do
      transition [:scheduled, :started] => :failed
    end

    event :import_retry do
      transition failed: :started
    end

    state :scheduled
    state :started
    state :finished
    state :failed

    after_transition [:none, :finished, :failed] => :scheduled do |project, _|
      project.run_after_commit do
        job_id = add_import_job
        update(import_jid: job_id) if job_id
      end
    end

    after_transition started: :finished do |project, _|
      project.reset_cache_and_import_attrs

      if Gitlab::ImportSources.importer_names.include?(project.import_type) && project.repo_exists?
        project.run_after_commit do
          Projects::AfterImportService.new(project).execute
        end
      end
    end
  end

  class << self
    # Searches for a list of projects based on the query given in `query`.
    #
    # On PostgreSQL this method uses "ILIKE" to perform a case-insensitive
    # search. On MySQL a regular "LIKE" is used as it's already
    # case-insensitive.
    #
    # query - The search query as a String.
    def search(query)
      ptable  = arel_table
      ntable  = Namespace.arel_table
      pattern = "%#{query}%"

      # unscoping unnecessary conditions that'll be applied
      # when executing `where("projects.id IN (#{union.to_sql})")`
      projects = unscoped.select(:id).where(
        ptable[:path].matches(pattern)
          .or(ptable[:name].matches(pattern))
          .or(ptable[:description].matches(pattern))
      )

      namespaces = unscoped.select(:id)
        .joins(:namespace)
        .where(ntable[:name].matches(pattern))

      union = Gitlab::SQL::Union.new([projects, namespaces])

      where("projects.id IN (#{union.to_sql})") # rubocop:disable GitlabSecurity/SqlInjection
    end

    def search_by_title(query)
      pattern = "%#{query}%"
      table   = Project.arel_table

      non_archived.where(table[:name].matches(pattern))
    end

    def visibility_levels
      Gitlab::VisibilityLevel.options
    end

    def sort(method)
      case method.to_s
      when 'storage_size_desc'
        # storage_size is a joined column so we need to
        # pass a string to avoid AR adding the table name
        reorder('project_statistics.storage_size DESC, projects.id DESC')
      when 'latest_activity_desc'
        reorder(last_activity_at: :desc)
      when 'latest_activity_asc'
        reorder(last_activity_at: :asc)
      else
        order_by(method)
      end
    end

    def reference_pattern
      %r{
        ((?<namespace>#{Gitlab::PathRegex::FULL_NAMESPACE_FORMAT_REGEX})\/)?
        (?<project>#{Gitlab::PathRegex::PROJECT_PATH_FORMAT_REGEX})
      }x
    end

    def trending
      joins('INNER JOIN trending_projects ON projects.id = trending_projects.project_id')
        .reorder('trending_projects.id ASC')
    end

    def cached_count
      Rails.cache.fetch('total_project_count', expires_in: 5.minutes) do
        Project.count
      end
    end

    def group_ids
      joins(:namespace).where(namespaces: { type: 'Group' }).select(:namespace_id)
    end
  end

  def lfs_enabled?
    return namespace.lfs_enabled? if self[:lfs_enabled].nil?

    self[:lfs_enabled] && Gitlab.config.lfs.enabled
  end

  def auto_devops_enabled?
    if auto_devops&.enabled.nil?
      current_application_settings.auto_devops_enabled?
    else
      auto_devops.enabled?
    end
  end

  def has_auto_devops_implicitly_disabled?
    auto_devops&.enabled.nil? && !current_application_settings.auto_devops_enabled?
  end

  def repository_storage_path
    Gitlab.config.repositories.storages[repository_storage].try(:[], 'path')
  end

  def team
    @team ||= ProjectTeam.new(self)
  end

  def repository
    @repository ||= Repository.new(full_path, self, disk_path: disk_path)
  end

  def reload_repository!
    @repository = nil
  end

  def container_registry_url
    if Gitlab.config.registry.enabled
      "#{Gitlab.config.registry.host_port}/#{full_path.downcase}"
    end
  end

  def has_container_registry_tags?
    return @images if defined?(@images)

    @images = container_repositories.to_a.any?(&:has_tags?) ||
      has_root_container_repository_tags?
  end

  def commit(ref = 'HEAD')
    repository.commit(ref)
  end

  # ref can't be HEAD, can only be branch/tag name or SHA
  def latest_successful_builds_for(ref = default_branch)
    latest_pipeline = pipelines.latest_successful_for(ref)

    if latest_pipeline
      latest_pipeline.builds.latest.with_artifacts
    else
      builds.none
    end
  end

  def merge_base_commit(first_commit_id, second_commit_id)
    sha = repository.merge_base(first_commit_id, second_commit_id)
    repository.commit(sha) if sha
  end

  def saved?
    id && persisted?
  end

  def add_import_job
    job_id =
      if forked?
        RepositoryForkWorker.perform_async(id,
                                           forked_from_project.repository_storage_path,
                                           forked_from_project.full_path,
                                           self.namespace.full_path)
      else
        RepositoryImportWorker.perform_async(self.id)
      end

    log_import_activity(job_id)

    job_id
  end

  def log_import_activity(job_id, type: :import)
    job_type = type.to_s.capitalize

    if job_id
      Rails.logger.info("#{job_type} job scheduled for #{full_path} with job ID #{job_id}.")
    else
      Rails.logger.error("#{job_type} job failed to create for #{full_path}.")
    end
  end

  def reset_cache_and_import_attrs
    run_after_commit do
      ProjectCacheWorker.perform_async(self.id)
    end

    update(import_error: nil)
    remove_import_data
  end

  # This method is overriden in EE::Project model
  def remove_import_data
    import_data&.destroy
  end

  def ci_config_path=(value)
    # Strip all leading slashes so that //foo -> foo
    super(value&.sub(%r{\A/+}, '')&.delete("\0"))
  end

  def import_url=(value)
    return super(value) unless Gitlab::UrlSanitizer.valid?(value)

    import_url = Gitlab::UrlSanitizer.new(value)
    super(import_url.sanitized_url)
    create_or_update_import_data(credentials: import_url.credentials)
  end

  def import_url
    if import_data && super.present?
      import_url = Gitlab::UrlSanitizer.new(super, credentials: import_data.credentials)
      import_url.full_url
    else
      super
    end
  end

  def valid_import_url?
    valid?(:import_url) || errors.messages[:import_url].nil?
  end

  def create_or_update_import_data(data: nil, credentials: nil)
    return unless import_url.present? && valid_import_url?

    project_import_data = import_data || build_import_data
    if data
      project_import_data.data ||= {}
      project_import_data.data = project_import_data.data.merge(data)
    end
    if credentials
      project_import_data.credentials ||= {}
      project_import_data.credentials = project_import_data.credentials.merge(credentials)
    end
  end

  def import?
    external_import? || forked? || gitlab_project_import?
  end

  def no_import?
    import_status == 'none'
  end

  def external_import?
    import_url.present?
  end

  def imported?
    import_finished?
  end

  def import_in_progress?
    import_started? || import_scheduled?
  end

  def import_started?
    import? && import_status == 'started'
  end

  def import_scheduled?
    import_status == 'scheduled'
  end

  def import_failed?
    import_status == 'failed'
  end

  def import_finished?
    import_status == 'finished'
  end

  def safe_import_url
    Gitlab::UrlSanitizer.new(import_url).masked_url
  end

  def gitlab_project_import?
    import_type == 'gitlab_project'
  end

  def gitea_import?
    import_type == 'gitea'
  end

  def github_import?
    import_type == 'github'
  end

  def check_limit
    unless creator.can_create_project? || namespace.kind == 'group'
      projects_limit = creator.projects_limit

      if projects_limit == 0
        self.errors.add(:limit_reached, "Personal project creation is not allowed. Please contact your administrator with questions")
      else
        self.errors.add(:limit_reached, "Your project limit is #{projects_limit} projects! Please contact your administrator to increase it")
      end
    end
  rescue
    self.errors.add(:base, "Can't check your ability to create project")
  end

  def visibility_level_allowed_by_group
    return if visibility_level_allowed_by_group?

    level_name = Gitlab::VisibilityLevel.level_name(self.visibility_level).downcase
    group_level_name = Gitlab::VisibilityLevel.level_name(self.group.visibility_level).downcase
    self.errors.add(:visibility_level, "#{level_name} is not allowed in a #{group_level_name} group.")
  end

  def visibility_level_allowed_as_fork
    return if visibility_level_allowed_as_fork?

    level_name = Gitlab::VisibilityLevel.level_name(self.visibility_level).downcase
    self.errors.add(:visibility_level, "#{level_name} is not allowed since the fork source project has lower visibility.")
  end

  def check_wiki_path_conflict
    return if path.blank?

    path_to_check = path.ends_with?('.wiki') ? path.chomp('.wiki') : "#{path}.wiki"

    if Project.where(namespace_id: namespace_id, path: path_to_check).exists?
      errors.add(:name, 'has already been taken')
    end
  end

  def to_param
    if persisted? && errors.include?(:path)
      path_was
    else
      path
    end
  end

  # `from` argument can be a Namespace or Project.
  def to_reference(from = nil, full: false)
    if full || cross_namespace_reference?(from)
      full_path
    elsif cross_project_reference?(from)
      path
    end
  end

  def to_human_reference(from_project = nil)
    if cross_namespace_reference?(from_project)
      name_with_namespace
    elsif cross_project_reference?(from_project)
      name
    end
  end

  def web_url
    Gitlab::Routing.url_helpers.project_url(self)
  end

  def new_issue_address(author)
    return unless Gitlab::IncomingEmail.supports_issue_creation? && author

    author.ensure_incoming_email_token!

    Gitlab::IncomingEmail.reply_address(
      "#{full_path}+#{author.incoming_email_token}")
  end

  def build_commit_note(commit)
    notes.new(commit_id: commit.id, noteable_type: 'Commit')
  end

  def last_activity
    last_event
  end

  def last_activity_date
    last_repository_updated_at || last_activity_at || updated_at
  end

  def project_id
    self.id
  end

  def get_issue(issue_id, current_user)
    issue = IssuesFinder.new(current_user, project_id: id).find_by(iid: issue_id) if issues_enabled?

    if issue
      issue
    elsif external_issue_tracker
      ExternalIssue.new(issue_id, self)
    end
  end

  def issue_exists?(issue_id)
    get_issue(issue_id)
  end

  def default_issue_tracker
    gitlab_issue_tracker_service || create_gitlab_issue_tracker_service
  end

  def issues_tracker
    if external_issue_tracker
      external_issue_tracker
    else
      default_issue_tracker
    end
  end

  def external_issue_reference_pattern
    external_issue_tracker.class.reference_pattern(only_long: issues_enabled?)
  end

  def default_issues_tracker?
    !external_issue_tracker
  end

  def external_issue_tracker
    if has_external_issue_tracker.nil? # To populate existing projects
      cache_has_external_issue_tracker
    end

    if has_external_issue_tracker?
      return @external_issue_tracker if defined?(@external_issue_tracker)

      @external_issue_tracker = services.external_issue_trackers.first
    else
      nil
    end
  end

  def cache_has_external_issue_tracker
    update_column(:has_external_issue_tracker, services.external_issue_trackers.any?) if Gitlab::Database.read_write?
  end

  def has_wiki?
    wiki_enabled? || has_external_wiki?
  end

  def external_wiki
    if has_external_wiki.nil?
      cache_has_external_wiki # Populate
    end

    if has_external_wiki
      @external_wiki ||= services.external_wikis.first
    else
      nil
    end
  end

  def cache_has_external_wiki
    update_column(:has_external_wiki, services.external_wikis.any?) if Gitlab::Database.read_write?
  end

  def find_or_initialize_services(exceptions: [])
    services_templates = Service.where(template: true)

    available_services_names = Service.available_services_names - exceptions

    available_services_names.map do |service_name|
      service = find_service(services, service_name)

      if service
        service
      else
        # We should check if template for the service exists
        template = find_service(services_templates, service_name)

        if template.nil?
          # If no template, we should create an instance. Ex `build_gitlab_ci_service`
          public_send("build_#{service_name}_service") # rubocop:disable GitlabSecurity/PublicSend
        else
          Service.build_from_template(id, template)
        end
      end
    end
  end

  def find_or_initialize_service(name)
    find_or_initialize_services.find { |service| service.to_param == name }
  end

  def create_labels
    Label.templates.each do |label|
      params = label.attributes.except('id', 'template', 'created_at', 'updated_at')
      Labels::FindOrCreateService.new(nil, self, params).execute(skip_authorization: true)
    end
  end

  def find_service(list, name)
    list.find { |service| service.to_param == name }
  end

  def ci_services
    services.where(category: :ci)
  end

  def ci_service
    @ci_service ||= ci_services.reorder(nil).find_by(active: true)
  end

  def deployment_services
    services.where(category: :deployment)
  end

  def deployment_service
    @deployment_service ||= deployment_services.reorder(nil).find_by(active: true)
  end

  def monitoring_services
    services.where(category: :monitoring)
  end

  def monitoring_service
    @monitoring_service ||= monitoring_services.reorder(nil).find_by(active: true)
  end

  def jira_tracker?
    issues_tracker.to_param == 'jira'
  end

  def avatar_type
    unless self.avatar.image?
      self.errors.add :avatar, 'only images allowed'
    end
  end

  def avatar_in_git
    repository.avatar
  end

  def avatar_url(**args)
    # We use avatar_path instead of overriding avatar_url because of carrierwave.
    # See https://gitlab.com/gitlab-org/gitlab-ce/merge_requests/11001/diffs#note_28659864
    avatar_path(args) || (Gitlab::Routing.url_helpers.project_avatar_url(self) if avatar_in_git)
  end

  # For compatibility with old code
  def code
    path
  end

  def items_for(entity)
    case entity
    when 'issue' then
      issues
    when 'merge_request' then
      merge_requests
    end
  end

  def send_move_instructions(old_path_with_namespace)
    # New project path needs to be committed to the DB or notification will
    # retrieve stale information
    run_after_commit { NotificationService.new.project_was_moved(self, old_path_with_namespace) }
  end

  def owner
    if group
      group
    else
      namespace.try(:owner)
    end
  end

  def execute_hooks(data, hooks_scope = :push_hooks)
    hooks.public_send(hooks_scope).each do |hook| # rubocop:disable GitlabSecurity/PublicSend
      hook.async_execute(data, hooks_scope.to_s)
    end
  end

  def execute_services(data, hooks_scope = :push_hooks)
    # Call only service hooks that are active for this scope
    services.public_send(hooks_scope).each do |service| # rubocop:disable GitlabSecurity/PublicSend
      service.async_execute(data)
    end
  end

  def valid_repo?
    repository.exists?
  rescue
    errors.add(:path, 'Invalid repository path')
    false
  end

  def repo
    repository.rugged
  end

  def url_to_repo
    gitlab_shell.url_to_repo(full_path)
  end

  def repo_exists?
    @repo_exists ||= repository.exists?
  rescue
    @repo_exists = false
  end

  def root_ref?(branch)
    repository.root_ref == branch
  end

  def ssh_url_to_repo
    url_to_repo
  end

  def http_url_to_repo
    "#{web_url}.git"
  end

  def user_can_push_to_empty_repo?(user)
    !ProtectedBranch.default_branch_protected? || team.max_member_access(user.id) > Gitlab::Access::DEVELOPER
  end

  def forked?
    return true if fork_network && fork_network.root_project != self

    # TODO: Use only the above conditional using the `fork_network`
    # This is the old conditional that looks at the `forked_project_link`, we
    # fall back to this while we're migrating the new models
    !(forked_project_link.nil? || forked_project_link.forked_from_project.nil?)
  end

  def personal?
    !group
  end

  # Expires various caches before a project is renamed.
  def expire_caches_before_rename(old_path)
    repo = Repository.new(old_path, self)
    wiki = Repository.new("#{old_path}.wiki", self)

    if repo.exists?
      repo.before_delete
    end

    if wiki.exists?
      wiki.before_delete
    end
  end

  # Check if repository already exists on disk
  def check_repository_path_availability
    return true if skip_disk_validation
    return false unless repository_storage_path

    expires_full_path_cache # we need to clear cache to validate renames correctly

    # Check if repository with same path already exists on disk we can
    # skip this for the hashed storage because the path does not change
    if legacy_storage? && repository_with_same_path_already_exists?
      errors.add(:base, 'There is already a repository with that name on disk')
      return false
    end

    true
  rescue GRPC::Internal # if the path is too long
    false
  end

  def create_repository(force: false)
    # Forked import is handled asynchronously
    return if forked? && !force

    if gitlab_shell.add_repository(repository_storage, disk_path)
      repository.after_create
      true
    else
      errors.add(:base, 'Failed to create repository via gitlab-shell')
      false
    end
  end

  def hook_attrs(backward: true)
    attrs = {
      name: name,
      description: description,
      web_url: web_url,
      avatar_url: avatar_url(only_path: false),
      git_ssh_url: ssh_url_to_repo,
      git_http_url: http_url_to_repo,
      namespace: namespace.name,
      visibility_level: visibility_level,
      path_with_namespace: full_path,
      default_branch: default_branch,
      ci_config_path: ci_config_path
    }

    # Backward compatibility
    if backward
      attrs.merge!({
                    homepage: web_url,
                    url: url_to_repo,
                    ssh_url: ssh_url_to_repo,
                    http_url: http_url_to_repo
                  })
    end

    attrs
  end

  def project_member(user)
    project_members.find_by(user_id: user)
  end

  def default_branch
    @default_branch ||= repository.root_ref if repository.exists?
  end

  def reload_default_branch
    @default_branch = nil
    default_branch
  end

  def visibility_level_field
    :visibility_level
  end

  def archive!
    update_attribute(:archived, true)
  end

  def unarchive!
    update_attribute(:archived, false)
  end

  def change_head(branch)
    if repository.branch_exists?(branch)
      repository.before_change_head
      repository.write_ref('HEAD', "refs/heads/#{branch}")
      repository.copy_gitattributes(branch)
      repository.after_change_head
      reload_default_branch
    else
      errors.add(:base, "Could not change HEAD: branch '#{branch}' does not exist")
      false
    end
  end

  def forked_from?(other_project)
    forked? && forked_from_project == other_project
  end

  def in_fork_network_of?(other_project)
    # TODO: Remove this in a next release when all fork_networks are populated
    # This makes sure all MergeRequests remain valid while the projects don't
    # have a fork_network yet.
    return true if forked_from?(other_project)

    return false if fork_network.nil? || other_project.fork_network.nil?

    fork_network == other_project.fork_network
  end

  def origin_merge_requests
    merge_requests.where(source_project_id: self.id)
  end

  def ensure_repository
    create_repository(force: true) unless repository_exists?
  end

  def repository_exists?
    !!repository.exists?
  end

  # update visibility_level of forks
  def update_forks_visibility_level
    return unless visibility_level < visibility_level_was

    forks.each do |forked_project|
      if forked_project.visibility_level > visibility_level
        forked_project.visibility_level = visibility_level
        forked_project.save!
      end
    end
  end

  def create_wiki
    ProjectWiki.new(self, self.owner).wiki
    true
  rescue ProjectWiki::CouldNotCreateWikiError
    errors.add(:base, 'Failed create wiki')
    false
  end

  def wiki
    @wiki ||= ProjectWiki.new(self, self.owner)
  end

  def jira_tracker_active?
    jira_tracker? && jira_service.active
  end

  def allowed_to_share_with_group?
    !namespace.share_with_group_lock
  end

  def pipeline_for(ref, sha = nil)
    sha ||= commit(ref).try(:sha)

    return unless sha

    pipelines.order(id: :desc).find_by(sha: sha, ref: ref)
  end

  def latest_successful_pipeline_for_default_branch
    if defined?(@latest_successful_pipeline_for_default_branch)
      return @latest_successful_pipeline_for_default_branch
    end

    @latest_successful_pipeline_for_default_branch =
      pipelines.latest_successful_for(default_branch)
  end

  def latest_successful_pipeline_for(ref = nil)
    if ref && ref != default_branch
      pipelines.latest_successful_for(ref)
    else
      latest_successful_pipeline_for_default_branch
    end
  end

  def enable_ci
    project_feature.update_attribute(:builds_access_level, ProjectFeature::ENABLED)
  end

  def shared_runners_available?
    shared_runners_enabled?
  end

  def shared_runners
    @shared_runners ||= shared_runners_available? ? Ci::Runner.shared : Ci::Runner.none
  end

  def active_shared_runners
    @active_shared_runners ||= shared_runners.active
  end

  def any_runners?(&block)
    active_runners.any?(&block) || active_shared_runners.any?(&block)
  end

  def valid_runners_token?(token)
    self.runners_token && ActiveSupport::SecurityUtils.variable_size_secure_compare(token, self.runners_token)
  end

  def build_timeout_in_minutes
    build_timeout / 60
  end

  def build_timeout_in_minutes=(value)
    self.build_timeout = value.to_i * 60
  end

  def open_issues_count
    Projects::OpenIssuesCountService.new(self).count
  end

  def open_merge_requests_count
    Projects::OpenMergeRequestsCountService.new(self).count
  end

  def visibility_level_allowed_as_fork?(level = self.visibility_level)
    return true unless forked?

    # self.forked_from_project will be nil before the project is saved, so
    # we need to go through the relation
    original_project = forked_project_link.forked_from_project
    return true unless original_project

    level <= original_project.visibility_level
  end

  def visibility_level_allowed_by_group?(level = self.visibility_level)
    return true unless group

    level <= group.visibility_level
  end

  def visibility_level_allowed?(level = self.visibility_level)
    visibility_level_allowed_as_fork?(level) && visibility_level_allowed_by_group?(level)
  end

  def runners_token
    ensure_runners_token!
  end

  def pages_deployed?
    Dir.exist?(public_pages_path)
  end

  def pages_url
    subdomain, _, url_path = full_path.partition('/')

    # The hostname always needs to be in downcased
    # All web servers convert hostname to lowercase
    host = "#{subdomain}.#{Settings.pages.host}".downcase

    # The host in URL always needs to be downcased
    url = Gitlab.config.pages.url.sub(/^https?:\/\//) do |prefix|
      "#{prefix}#{subdomain}."
    end.downcase

    # If the project path is the same as host, we serve it as group page
    return url if host == url_path

    "#{url}/#{url_path}"
  end

  def pages_subdomain
    full_path.partition('/').first
  end

  def pages_path
    # TODO: when we migrate Pages to work with new storage types, change here to use disk_path
    File.join(Settings.pages.path, full_path)
  end

  def public_pages_path
    File.join(pages_path, 'public')
  end

  def pages_available?
    Gitlab.config.pages.enabled && !namespace.subgroup?
  end

  def remove_private_deploy_keys
    exclude_keys_linked_to_other_projects = <<-SQL
      NOT EXISTS (
        SELECT 1
        FROM deploy_keys_projects dkp2
        WHERE dkp2.deploy_key_id = deploy_keys_projects.deploy_key_id
        AND dkp2.project_id != deploy_keys_projects.project_id
      )
    SQL

    deploy_keys.where(public: false)
               .where(exclude_keys_linked_to_other_projects)
               .delete_all
  end

  # TODO: what to do here when not using Legacy Storage? Do we still need to rename and delay removal?
  def remove_pages
    # Projects with a missing namespace cannot have their pages removed
    return unless namespace

    ::Projects::UpdatePagesConfigurationService.new(self).execute

    # 1. We rename pages to temporary directory
    # 2. We wait 5 minutes, due to NFS caching
    # 3. We asynchronously remove pages with force
    temp_path = "#{path}.#{SecureRandom.hex}.deleted"

    if Gitlab::PagesTransfer.new.rename_project(path, temp_path, namespace.full_path)
      PagesWorker.perform_in(5.minutes, :remove, namespace.full_path, temp_path)
    end
  end

  def rename_repo
    new_full_path = build_full_path

    Rails.logger.error "Attempting to rename #{full_path_was} -> #{new_full_path}"

    if has_container_registry_tags?
      Rails.logger.error "Project #{full_path_was} cannot be renamed because container registry tags are present!"

      # we currently doesn't support renaming repository if it contains images in container registry
      raise StandardError.new('Project cannot be renamed, because images are present in its container registry')
    end

    expire_caches_before_rename(full_path_was)

    if storage.rename_repo
      Gitlab::AppLogger.info "Project was renamed: #{full_path_was} -> #{new_full_path}"
      rename_repo_notify!
      after_rename_repo
    else
      Rails.logger.error "Repository could not be renamed: #{full_path_was} -> #{new_full_path}"

      # if we cannot move namespace directory we should rollback
      # db changes in order to prevent out of sync between db and fs
      raise StandardError.new('repository cannot be renamed')
    end
  end

  def rename_repo_notify!
    send_move_instructions(full_path_was)
    expires_full_path_cache

    self.old_path_with_namespace = full_path_was
    SystemHooksService.new.execute_hooks_for(self, :rename)

    reload_repository!
  end

  def after_rename_repo
    path_before_change = previous_changes['path'].first

    Gitlab::UploadsTransfer.new.rename_project(path_before_change, self.path, namespace.full_path)
    Gitlab::PagesTransfer.new.rename_project(path_before_change, self.path, namespace.full_path)
  end

  def running_or_pending_build_count(force: false)
    Rails.cache.fetch(['projects', id, 'running_or_pending_build_count'], force: force) do
      builds.running_or_pending.count(:all)
    end
  end

  # Lazy loading of the `pipeline_status` attribute
  def pipeline_status
    @pipeline_status ||= Gitlab::Cache::Ci::ProjectPipelineStatus.load_for_project(self)
  end

  def mark_import_as_failed(error_message)
    original_errors = errors.dup
    sanitized_message = Gitlab::UrlSanitizer.sanitize(error_message)

    import_fail
    update_column(:import_error, sanitized_message)
  rescue ActiveRecord::ActiveRecordError => e
    Rails.logger.error("Error setting import status to failed: #{e.message}. Original error: #{sanitized_message}")
  ensure
    @errors = original_errors
  end

  def add_export_job(current_user:)
    job_id = ProjectExportWorker.perform_async(current_user.id, self.id)

    if job_id
      Rails.logger.info "Export job started for project ID #{self.id} with job ID #{job_id}"
    else
      Rails.logger.error "Export job failed to start for project ID #{self.id}"
    end
  end

  def export_path
    File.join(Gitlab::ImportExport.storage_path, disk_path)
  end

  def export_project_path
    Dir.glob("#{export_path}/*export.tar.gz").max_by { |f| File.ctime(f) }
  end

  def remove_exports
    _, status = Gitlab::Popen.popen(%W(find #{export_path} -not -path #{export_path} -delete))
    status.zero?
  end

  def full_path_slug
    Gitlab::Utils.slugify(full_path.to_s)
  end

  def has_ci?
    repository.gitlab_ci_yml || auto_devops_enabled?
  end

  def predefined_variables
    [
      { key: 'CI_PROJECT_ID', value: id.to_s, public: true },
      { key: 'CI_PROJECT_NAME', value: path, public: true },
      { key: 'CI_PROJECT_PATH', value: full_path, public: true },
      { key: 'CI_PROJECT_PATH_SLUG', value: full_path_slug, public: true },
      { key: 'CI_PROJECT_NAMESPACE', value: namespace.full_path, public: true },
      { key: 'CI_PROJECT_URL', value: web_url, public: true }
    ]
  end

  def container_registry_variables
    return [] unless Gitlab.config.registry.enabled

    variables = [
      { key: 'CI_REGISTRY', value: Gitlab.config.registry.host_port, public: true }
    ]

    if container_registry_enabled?
      variables << { key: 'CI_REGISTRY_IMAGE', value: container_registry_url, public: true }
    end

    variables
  end

  def secret_variables_for(ref:, environment: nil)
    # EE would use the environment
    if protected_for?(ref)
      variables
    else
      variables.unprotected
    end
  end

  def protected_for?(ref)
    ProtectedBranch.protected?(self, ref) ||
      ProtectedTag.protected?(self, ref)
  end

  def deployment_variables
    return [] unless deployment_service

    deployment_service.predefined_variables
  end

  def auto_devops_variables
    return [] unless auto_devops_enabled?

    auto_devops&.variables || []
  end

  def append_or_update_attribute(name, value)
    old_values = public_send(name.to_s) # rubocop:disable GitlabSecurity/PublicSend

    if Project.reflect_on_association(name).try(:macro) == :has_many && old_values.any?
      update_attribute(name, old_values + value)
    else
      update_attribute(name, value)
    end

  rescue ActiveRecord::RecordNotSaved => e
    handle_update_attribute_error(e, value)
  end

  def pushes_since_gc
    Gitlab::Redis::SharedState.with { |redis| redis.get(pushes_since_gc_redis_shared_state_key).to_i }
  end

  def increment_pushes_since_gc
    Gitlab::Redis::SharedState.with { |redis| redis.incr(pushes_since_gc_redis_shared_state_key) }
  end

  def reset_pushes_since_gc
    Gitlab::Redis::SharedState.with { |redis| redis.del(pushes_since_gc_redis_shared_state_key) }
  end

  def route_map_for(commit_sha)
    @route_maps_by_commit ||= Hash.new do |h, sha|
      h[sha] = begin
        data = repository.route_map_for(sha)
        next unless data

        Gitlab::RouteMap.new(data)
      rescue Gitlab::RouteMap::FormatError
        nil
      end
    end

    @route_maps_by_commit[commit_sha]
  end

  def public_path_for_source_path(path, commit_sha)
    map = route_map_for(commit_sha)
    return unless map

    map.public_path_for_source_path(path)
  end

  def parent
    namespace
  end

  def parent_changed?
    namespace_id_changed?
  end

  def default_merge_request_target
    if forked_from_project&.merge_requests_enabled?
      forked_from_project
    else
      self
    end
  end

  def multiple_issue_boards_available?(user)
    feature_available?(:multiple_issue_boards, user)
  end

  def issue_board_milestone_available?(user = nil)
    feature_available?(:issue_board_milestone, user)
  end

  def full_path_was
    File.join(namespace.full_path, previous_changes['path'].first)
  end

  alias_method :name_with_namespace, :full_name
  alias_method :human_name, :full_name
  # @deprecated cannot remove yet because it has an index with its name in elasticsearch
  alias_method :path_with_namespace, :full_path

  def forks_count
    Projects::ForksCountService.new(self).count
  end

  def legacy_storage?
    [nil, 0].include?(self.storage_version)
  end

  def hashed_storage?
    self.storage_version && self.storage_version >= 1
  end

  def renamed?
    persisted? && path_changed?
  end

  def merge_method
    if self.merge_requests_ff_only_enabled
      :ff
    elsif self.merge_requests_rebase_enabled
      :rebase_merge
    else
      :merge
    end
  end

  def merge_method=(method)
    case method.to_s
    when "ff"
      self.merge_requests_ff_only_enabled = true
      self.merge_requests_rebase_enabled = true
    when "rebase_merge"
      self.merge_requests_ff_only_enabled = false
      self.merge_requests_rebase_enabled = true
    when "merge"
      self.merge_requests_ff_only_enabled = false
      self.merge_requests_rebase_enabled = false
    end
  end

  def ff_merge_must_be_possible?
    self.merge_requests_ff_only_enabled || self.merge_requests_rebase_enabled
  end

  def migrate_to_hashed_storage!
    return if hashed_storage?

    update!(repository_read_only: true)

    if repo_reference_count > 0 || wiki_reference_count > 0
      ProjectMigrateHashedStorageWorker.perform_in(Gitlab::ReferenceCounter::REFERENCE_EXPIRE_TIME, id)
    else
      ProjectMigrateHashedStorageWorker.perform_async(id)
    end
  end

  def storage_version=(value)
    super

    @storage = nil if storage_version_changed?
  end

  def gl_repository(is_wiki:)
    Gitlab::GlRepository.gl_repository(self, is_wiki)
  end

  private

  def storage
    @storage ||=
      if hashed_storage?
        Storage::HashedProject.new(self)
      else
        Storage::LegacyProject.new(self)
      end
  end

  def use_hashed_storage
    if self.new_record? && current_application_settings.hashed_storage_enabled
      self.storage_version = LATEST_STORAGE_VERSION
    end
  end

  def repo_reference_count
    Gitlab::ReferenceCounter.new(gl_repository(is_wiki: false)).value
  end

  def wiki_reference_count
    Gitlab::ReferenceCounter.new(gl_repository(is_wiki: true)).value
  end

  def check_repository_absence!
    return if skip_disk_validation

    if repository_storage_path.blank? || repository_with_same_path_already_exists?
      errors.add(:base, 'There is already a repository with that name on disk')
      throw :abort
    end
  end

  def repository_with_same_path_already_exists?
    gitlab_shell.exists?(repository_storage_path, "#{disk_path}.git")
  end

  # set last_activity_at to the same as created_at
  def set_last_activity_at
    update_column(:last_activity_at, self.created_at)
  end

  def set_last_repository_updated_at
    update_column(:last_repository_updated_at, self.created_at)
  end

  def cross_namespace_reference?(from)
    case from
    when Project
      namespace != from.namespace
    when Namespace
      namespace != from
    end
  end

  # Check if a reference is being done cross-project
  def cross_project_reference?(from)
    return true if from.is_a?(Namespace)

    from && self != from
  end

  def pushes_since_gc_redis_shared_state_key
    "projects/#{id}/pushes_since_gc"
  end

  # Similar to the normal callbacks that hook into the life cycle of an
  # Active Record object, you can also define callbacks that get triggered
  # when you add an object to an association collection. If any of these
  # callbacks throw an exception, the object will not be added to the
  # collection. Before you add a new board to the boards collection if you
  # already have 1, 2, or n it will fail, but it if you have 0 that is lower
  # than the number of permitted boards per project it won't fail.
  def validate_board_limit(board)
    raise BoardLimitExceeded, 'Number of permitted boards exceeded' if boards.size >= NUMBER_OF_PERMITTED_BOARDS
  end

  def update_project_statistics
    stats = statistics || build_statistics
    stats.update(namespace_id: namespace_id)
  end

  def check_pending_delete
    return if valid_attribute?(:name) && valid_attribute?(:path)
    return unless pending_delete_twin

    %i[route route.path name path].each do |error|
      errors.delete(error)
    end

    errors.add(:base, "The project is still being deleted. Please try again later.")
  end

  def pending_delete_twin
    return false unless path

    Project.pending_delete.find_by_full_path(full_path)
  end

  ##
  # This method is here because of support for legacy container repository
  # which has exactly the same path like project does, but which might not be
  # persisted in `container_repositories` table.
  #
  def has_root_container_repository_tags?
    return false unless Gitlab.config.registry.enabled

    ContainerRepository.build_root_repository(self).has_tags?
  end

  def handle_update_attribute_error(ex, value)
    if ex.message.start_with?('Failed to replace')
      if value.respond_to?(:each)
        invalid = value.detect(&:invalid?)

        raise ex, ([ex.message] + invalid.errors.full_messages).join(' ') if invalid
      end
    end

    raise ex
  end
end