clara-temporalio 0.4.3-x86_64-darwin

data/lib/temporalio/internal/worker/workflow_instance.rb

@@ -0,0 +1,800 @@
+# frozen_string_literal: true
+
+require 'json'
+require 'temporalio'
+require 'temporalio/activity/definition'
+require 'temporalio/api'
+require 'temporalio/converters/payload_converter'
+require 'temporalio/converters/raw_value'
+require 'temporalio/error'
+require 'temporalio/internal/bridge/api'
+require 'temporalio/internal/proto_utils'
+require 'temporalio/internal/worker/workflow_instance/child_workflow_handle'
+require 'temporalio/internal/worker/workflow_instance/context'
+require 'temporalio/internal/worker/workflow_instance/details'
+require 'temporalio/internal/worker/workflow_instance/externally_immutable_hash'
+require 'temporalio/internal/worker/workflow_instance/handler_execution'
+require 'temporalio/internal/worker/workflow_instance/handler_hash'
+require 'temporalio/internal/worker/workflow_instance/illegal_call_tracer'
+require 'temporalio/internal/worker/workflow_instance/inbound_implementation'
+require 'temporalio/internal/worker/workflow_instance/outbound_implementation'
+require 'temporalio/internal/worker/workflow_instance/replay_safe_logger'
+require 'temporalio/internal/worker/workflow_instance/replay_safe_metric'
+require 'temporalio/internal/worker/workflow_instance/scheduler'
+require 'temporalio/retry_policy'
+require 'temporalio/scoped_logger'
+require 'temporalio/worker/interceptor'
+require 'temporalio/worker_deployment_version'
+require 'temporalio/workflow/info'
+require 'temporalio/workflow/update_info'
+require 'timeout'
+
+module Temporalio
+  module Internal
+    module Worker
+      # Instance of a user workflow. This is the instance with all state needed to run the workflow and is expected to
+      # be cached by the worker for sticky execution.
+      class WorkflowInstance
+        def self.new_completion_with_failure(run_id:, error:, failure_converter:, payload_converter:)
+          Bridge::Api::WorkflowCompletion::WorkflowActivationCompletion.new(
+            run_id: run_id,
+            failed: Bridge::Api::WorkflowCompletion::Failure.new(
+              failure: begin
+                failure_converter.to_failure(error, payload_converter)
+              rescue Exception => e # rubocop:disable Lint/RescueException
+                Api::Failure::V1::Failure.new(
+                  message: "Failed converting error to failure: #{e.message}, " \
+                           "original error message: #{error.message}",
+                  application_failure_info: Api::Failure::V1::ApplicationFailureInfo.new
+                )
+              end
+            )
+          )
+        end
+
+        attr_reader :context, :logger, :info, :scheduler, :disable_eager_activity_execution, :pending_activities,
+                    :pending_timers, :pending_child_workflow_starts, :pending_child_workflows,
+                    :pending_external_signals, :pending_external_cancels, :in_progress_handlers, :payload_converter,
+                    :failure_converter, :cancellation, :continue_as_new_suggested, :current_deployment_version,
+                    :current_history_length, :current_history_size, :replaying, :random,
+                    :signal_handlers, :query_handlers, :update_handlers, :context_frozen, :assert_valid_local_activity
+        attr_accessor :io_enabled, :current_details
+
+        def initialize(details)
+          # Initialize general state
+          @context = Context.new(self)
+          if details.illegal_calls && !details.illegal_calls.empty?
+            @tracer = IllegalCallTracer.new(details.illegal_calls)
+          end
+          @logger = ReplaySafeLogger.new(logger: details.logger, instance: self)
+          @logger.scoped_values_getter = proc { scoped_logger_info }
+          @runtime_metric_meter = details.metric_meter
+          @io_enabled = details.unsafe_workflow_io_enabled
+          @scheduler = Scheduler.new(self)
+          @payload_converter = details.payload_converter
+          @failure_converter = details.failure_converter
+          @disable_eager_activity_execution = details.disable_eager_activity_execution
+          @pending_activities = {} # Keyed by sequence, value is fiber to resume with proto result
+          @pending_timers = {} # Keyed by sequence, value is fiber to resume with proto result
+          @pending_child_workflow_starts = {} # Keyed by sequence, value is fiber to resume with proto result
+          @pending_child_workflows = {} # Keyed by sequence, value is ChildWorkflowHandle to resolve with proto result
+          @pending_external_signals = {} # Keyed by sequence, value is fiber to resume with proto result
+          @pending_external_cancels = {} # Keyed by sequence, value is fiber to resume with proto result
+          @buffered_signals = {} # Keyed by signal name, value is array of signal jobs
+          # TODO(cretz): Should these be sets instead? Both should be fairly low counts.
+          @in_progress_handlers = [] # Value is HandlerExecution
+          @patches_notified = []
+          @definition = details.definition
+          @interceptors = details.interceptors
+          @cancellation, @cancellation_proc = Cancellation.new
+          @continue_as_new_suggested = false
+          @current_history_length = 0
+          @current_history_size = 0
+          @replaying = false
+          @workflow_failure_exception_types = details.workflow_failure_exception_types
+          @signal_handlers = HandlerHash.new(
+            details.definition.signals,
+            Workflow::Definition::Signal
+          ) do |defn|
+            # New definition, drain buffer. If it's dynamic (i.e. no name) drain them all.
+            to_drain = if defn.name.nil?
+                         all_signals = @buffered_signals.values.flatten
+                         @buffered_signals.clear
+                         all_signals
+                       else
+                         @buffered_signals.delete(defn.name)
+                       end
+            to_drain&.each { |job| apply_signal(job) }
+          end
+          @query_handlers = HandlerHash.new(details.definition.queries, Workflow::Definition::Query)
+          @update_handlers = HandlerHash.new(details.definition.updates, Workflow::Definition::Update)
+          @definition_options = Workflow::DefinitionOptions.new(
+            failure_exception_types: details.definition.failure_exception_types,
+            versioning_behavior: details.definition.versioning_behavior
+          )
+
+          @assert_valid_local_activity = details.assert_valid_local_activity
+
+          # Create all things needed from initial job
+          @init_job = details.initial_activation.jobs.find { |j| !j.initialize_workflow.nil? }&.initialize_workflow
+          raise 'Missing init job from first activation' unless @init_job
+
+          illegal_call_tracing_disabled do
+            @info = Workflow::Info.new(
+              attempt: @init_job.attempt,
+              continued_run_id: ProtoUtils.string_or(@init_job.continued_from_execution_run_id),
+              cron_schedule: ProtoUtils.string_or(@init_job.cron_schedule),
+              execution_timeout: ProtoUtils.duration_to_seconds(@init_job.workflow_execution_timeout),
+              headers: ProtoUtils.headers_from_proto_map(@init_job.headers, @payload_converter) || {},
+              last_failure: if @init_job.continued_failure
+                              @failure_converter.from_failure(@init_job.continued_failure, @payload_converter)
+                            end,
+              last_result: if @init_job.last_completion_result
+                             @payload_converter.from_payloads(@init_job.last_completion_result).first
+                           end,
+              namespace: details.namespace,
+              parent: if @init_job.parent_workflow_info
+                        Workflow::Info::ParentInfo.new(
+                          namespace: @init_job.parent_workflow_info.namespace,
+                          run_id: @init_job.parent_workflow_info.run_id,
+                          workflow_id: @init_job.parent_workflow_info.workflow_id
+                        )
+                      end,
+              priority: Priority._from_proto(@init_job.priority),
+              retry_policy: (RetryPolicy._from_proto(@init_job.retry_policy) if @init_job.retry_policy),
+              root: if @init_job.root_workflow
+                      Workflow::Info::RootInfo.new(
+                        run_id: @init_job.root_workflow.run_id,
+                        workflow_id: @init_job.root_workflow.workflow_id
+                      )
+                    end,
+              run_id: details.initial_activation.run_id,
+              run_timeout: ProtoUtils.duration_to_seconds(@init_job.workflow_run_timeout),
+              start_time: ProtoUtils.timestamp_to_time(@init_job.start_time) || raise,
+              task_queue: details.task_queue,
+              task_timeout: ProtoUtils.duration_to_seconds(@init_job.workflow_task_timeout) || raise,
+              workflow_id: @init_job.workflow_id,
+              workflow_type: @init_job.workflow_type
+            ).freeze
+
+            @random = Random.new(@init_job.randomness_seed)
+          end
+        end
+
+        def activate(activation)
+          # Run inside of scheduler
+          run_in_scheduler { activate_internal(activation) }
+        end
+
+        def add_command(command)
+          raise Workflow::InvalidWorkflowStateError, 'Cannot add commands in this context' if @context_frozen
+
+          @commands << command
+        end
+
+        def instance
+          @instance or raise 'Instance accessed before created'
+        end
+
+        def search_attributes
+          # Lazy on first access
+          @search_attributes ||= SearchAttributes._from_proto(
+            @init_job.search_attributes, disable_mutations: true, never_nil: true
+          ) || raise
+        end
+
+        def memo
+          # Lazy on first access
+          @memo ||= ExternallyImmutableHash.new(ProtoUtils.memo_from_proto(@init_job.memo, payload_converter) || {})
+        end
+
+        def now
+          # Create each time
+          ProtoUtils.timestamp_to_time(@now_timestamp) or raise 'Time unexpectedly not present'
+        end
+
+        def illegal_call_tracing_disabled(&)
+          @tracer.disable(&)
+        end
+
+        def patch(patch_id:, deprecated:)
+          # Use memoized result if present. If this is being deprecated, we can still use memoized result and skip the
+          # command.
+          patch_id = patch_id.to_s
+          @patches_memoized ||= {}
+          @patches_memoized.fetch(patch_id) do
+            patched = !replaying || @patches_notified.include?(patch_id)
+            @patches_memoized[patch_id] = patched
+            if patched
+              add_command(
+                Bridge::Api::WorkflowCommands::WorkflowCommand.new(
+                  set_patch_marker: Bridge::Api::WorkflowCommands::SetPatchMarker.new(patch_id:, deprecated:)
+                )
+              )
+            end
+            patched
+          end
+        end
+
+        def metric_meter
+          @metric_meter ||= ReplaySafeMetric::Meter.new(
+            @runtime_metric_meter.with_additional_attributes(
+              {
+                namespace: info.namespace,
+                task_queue: info.task_queue,
+                workflow_type: info.workflow_type
+              }
+            )
+          )
+        end
+
+        private
+
+        def run_in_scheduler(&)
+          Fiber.set_scheduler(@scheduler)
+          if @tracer
+            @tracer.enable(&)
+          else
+            yield
+          end
+        ensure
+          Fiber.set_scheduler(nil)
+        end
+
+        def activate_internal(activation)
+          # Reset some activation state
+          @commands = []
+          @current_activation_error = nil
+          @continue_as_new_suggested = activation.continue_as_new_suggested
+          @current_deployment_version = WorkerDeploymentVersion._from_bridge(
+            activation.deployment_version_for_current_task
+          )
+          @current_history_length = activation.history_length
+          @current_history_size = activation.history_size_bytes
+          @replaying = activation.is_replaying
+          @now_timestamp = activation.timestamp
+
+          # Apply jobs and run event loop
+          begin
+            # Create instance if it doesn't already exist
+            @instance ||= with_context_frozen { create_instance }
+
+            # Apply jobs
+            activation.jobs.each { |job| apply(job) }
+
+            # Schedule primary 'execute' if not already running (i.e. this is
+            # the first activation)
+            @primary_fiber ||= schedule(top_level: true) { run_workflow }
+
+            # Run the event loop
+            @scheduler.run_until_all_yielded
+          rescue Exception => e # rubocop:disable Lint/RescueException
+            on_top_level_exception(e)
+          end
+
+          # If we are not replaying and workflow is complete but not a
+          # failure (i.e. success, continue as new, or cancel), we warn for
+          # any unfinished handlers.
+          if !@replaying && @commands.any? do |c|
+            !c.complete_workflow_execution.nil? ||
+            !c.continue_as_new_workflow_execution.nil? ||
+            !c.cancel_workflow_execution.nil?
+          end
+            warn_on_any_unfinished_handlers
+          end
+
+          # Return success or failure
+          if @current_activation_error
+            @logger.replay_safety_disabled do
+              @logger.warn('Failed activation')
+              @logger.warn(@current_activation_error)
+            end
+            WorkflowInstance.new_completion_with_failure(
+              run_id: activation.run_id,
+              error: @current_activation_error,
+              failure_converter: @failure_converter,
+              payload_converter: @payload_converter
+            )
+          else
+            Bridge::Api::WorkflowCompletion::WorkflowActivationCompletion.new(
+              run_id: activation.run_id,
+              successful: Bridge::Api::WorkflowCompletion::Success.new(
+                commands: @commands, versioning_behavior: @definition_options.versioning_behavior
+              )
+            )
+          end
+        ensure
+          @commands = nil
+          @current_activation_error = nil
+        end
+
+        def create_instance
+          # Convert workflow arguments
+          @workflow_arguments = convert_args(payload_array: @init_job.arguments,
+                                             method_name: :execute,
+                                             raw_args: @definition.raw_args)
+
+          # Initialize interceptors
+          @inbound = @interceptors.reverse_each.reduce(InboundImplementation.new(self)) do |acc, int|
+            int.intercept_workflow(acc)
+          end
+          @inbound.init(OutboundImplementation.new(self))
+
+          # Create the user instance
+          instance = if @definition.init
+                       @definition.workflow_class.new(*@workflow_arguments)
+                     else
+                       @definition.workflow_class.new
+                     end
+
+          # Run Dynamic config getter
+          if @definition.dynamic_options_method
+            dynamic_options = instance.send(@definition.dynamic_options_method)
+            if dynamic_options&.versioning_behavior
+              @definition_options.versioning_behavior = dynamic_options.versioning_behavior
+            end
+            if dynamic_options&.failure_exception_types
+              @definition_options.failure_exception_types = dynamic_options.failure_exception_types
+            end
+          end
+
+          instance
+        end
+
+        def apply(job)
+          case job.variant
+          when :initialize_workflow
+            # Ignore
+          when :fire_timer
+            pending_timers.delete(job.fire_timer.seq)&.resume
+          when :update_random_seed
+            @random = illegal_call_tracing_disabled { Random.new(job.update_random_seed.randomness_seed) }
+          when :query_workflow
+            apply_query(job.query_workflow)
+          when :cancel_workflow
+            # TODO(cretz): Use the details somehow?
+            @cancellation_proc.call(reason: 'Workflow canceled')
+          when :signal_workflow
+            apply_signal(job.signal_workflow)
+          when :resolve_activity
+            pending_activities.delete(job.resolve_activity.seq)&.resume(job.resolve_activity.result)
+          when :notify_has_patch
+            @patches_notified << job.notify_has_patch.patch_id
+          when :resolve_child_workflow_execution_start
+            pending_child_workflow_starts.delete(job.resolve_child_workflow_execution_start.seq)&.resume(
+              job.resolve_child_workflow_execution_start
+            )
+          when :resolve_child_workflow_execution
+            pending_child_workflows.delete(job.resolve_child_workflow_execution.seq)&._resolve(
+              job.resolve_child_workflow_execution.result
+            )
+          when :resolve_signal_external_workflow
+            pending_external_signals.delete(job.resolve_signal_external_workflow.seq)&.resume(
+              job.resolve_signal_external_workflow
+            )
+          when :resolve_request_cancel_external_workflow
+            pending_external_cancels.delete(job.resolve_request_cancel_external_workflow.seq)&.resume(
+              job.resolve_request_cancel_external_workflow
+            )
+          when :do_update
+            apply_update(job.do_update)
+          else
+            raise "Unrecognized activation job variant: #{job.variant}"
+          end
+        end
+
+        def apply_signal(job)
+          # Get signal definition, falling back to dynamic if not present and not reserved
+          defn = signal_handlers[job.signal_name]
+          defn = signal_handlers[nil] if !defn && !Internal::ProtoUtils.reserved_name?(job.signal_name)
+
+          handler_exec =
+            if defn
+              HandlerExecution.new(name: job.signal_name, update_id: nil, unfinished_policy: defn.unfinished_policy)
+            end
+          # Process as a top level handler so that errors are treated as if in primary workflow method
+          schedule(top_level: true, handler_exec:) do
+            # Send to interceptor if there is a definition, buffer otherwise
+            if defn
+              @inbound.handle_signal(
+                Temporalio::Worker::Interceptor::Workflow::HandleSignalInput.new(
+                  signal: job.signal_name,
+                  args: begin
+                    convert_handler_args(payload_array: job.input, defn:)
+                  rescue StandardError => e
+                    # Signals argument conversion failure must not fail task
+                    @logger.error("Failed converting signal input arguments for #{job.signal_name}, dropping signal")
+                    @logger.error(e)
+                    next
+                  end,
+                  definition: defn,
+                  headers: ProtoUtils.headers_from_proto_map(job.headers, @payload_converter) || {}
+                )
+              )
+            else
+              buffered = @buffered_signals[job.signal_name]
+              buffered = @buffered_signals[job.signal_name] = [] if buffered.nil?
+              buffered << job
+            end
+          end
+        end
+
+        def apply_query(job)
+          schedule do
+            # If it's a built-in, run it without interceptors, otherwise do normal behavior
+            result = if job.query_type == '__stack_trace'
+                       # Use raw value built from default converter because we don't want to use user-conversion
+                       Converters::RawValue.new(Converters::PayloadConverter.default.to_payload(scheduler.stack_trace))
+                     elsif job.query_type == '__temporal_workflow_metadata'
+                       # Use raw value built from default converter because we don't want to use user-conversion
+                       Converters::RawValue.new(Converters::PayloadConverter.default.to_payload(workflow_metadata))
+                     else
+                       # Get query definition, falling back to dynamic if not present and not reserved
+                       defn = query_handlers[job.query_type]
+                       defn = query_handlers[nil] if !defn && !Internal::ProtoUtils.reserved_name?(job.query_type)
+
+                       unless defn
+                         raise "Query handler for #{job.query_type} expected but not found, " \
+                               "known queries: [#{query_handlers.keys.compact.sort.join(', ')}]"
+                       end
+
+                       with_context_frozen do
+                         @inbound.handle_query(
+                           Temporalio::Worker::Interceptor::Workflow::HandleQueryInput.new(
+                             id: job.query_id,
+                             query: job.query_type,
+                             args: begin
+                               convert_handler_args(payload_array: job.arguments, defn:)
+                             rescue StandardError => e
+                               raise "Failed converting query input arguments: #{e}"
+                             end,
+                             definition: defn,
+                             headers: ProtoUtils.headers_from_proto_map(job.headers, @payload_converter) || {}
+                           )
+                         )
+                       end
+                     end
+
+            add_command(
+              Bridge::Api::WorkflowCommands::WorkflowCommand.new(
+                respond_to_query: Bridge::Api::WorkflowCommands::QueryResult.new(
+                  query_id: job.query_id,
+                  succeeded: Bridge::Api::WorkflowCommands::QuerySuccess.new(
+                    response: @payload_converter.to_payload(result)
+                  )
+                )
+              )
+            )
+          rescue Exception => e # rubocop:disable Lint/RescueException
+            add_command(
+              Bridge::Api::WorkflowCommands::WorkflowCommand.new(
+                respond_to_query: Bridge::Api::WorkflowCommands::QueryResult.new(
+                  query_id: job.query_id,
+                  failed: @failure_converter.to_failure(e, @payload_converter)
+                )
+              )
+            )
+          end
+        end
+
+        def apply_update(job)
+          # Get update definition, falling back to dynamic if not present and not reserved
+          defn = update_handlers[job.name]
+          defn = update_handlers[nil] if !defn && !Internal::ProtoUtils.reserved_name?(job.name)
+
+          handler_exec =
+            (HandlerExecution.new(name: job.name, update_id: job.id, unfinished_policy: defn.unfinished_policy) if defn)
+          schedule(handler_exec:) do
+            # Until this is accepted, all errors are rejections
+            accepted = false
+
+            # Set update info
+            Fiber[:__temporal_update_info] = Workflow::UpdateInfo.new(id: job.id, name: job.name).freeze
+
+            # Reject if not present
+            unless defn
+              raise "Update handler for #{job.name} expected but not found, " \
+                    "known updates: [#{update_handlers.keys.compact.sort.join(', ')}]"
+            end
+
+            # To match other SDKs, we are only calling the validation interceptor if there is a validator. Also to match
+            # other SDKs, we are re-converting the args between validate and update to disallow user mutation in
+            # validator/interceptor.
+            if job.run_validator && defn.validator_to_invoke
+              with_context_frozen do
+                @inbound.validate_update(
+                  Temporalio::Worker::Interceptor::Workflow::HandleUpdateInput.new(
+                    id: job.id,
+                    update: job.name,
+                    args: begin
+                      convert_handler_args(payload_array: job.input, defn:)
+                    rescue StandardError => e
+                      raise "Failed converting update input arguments: #{e}"
+                    end,
+                    definition: defn,
+                    headers: ProtoUtils.headers_from_proto_map(job.headers, @payload_converter) || {}
+                  )
+                )
+              end
+            end
+
+            # We build the input before marking accepted so the exception can reject instead of fail task
+            input = Temporalio::Worker::Interceptor::Workflow::HandleUpdateInput.new(
+              id: job.id,
+              update: job.name,
+              args: begin
+                convert_handler_args(payload_array: job.input, defn:)
+              rescue StandardError => e
+                raise "Failed converting update input arguments: #{e}"
+              end,
+              definition: defn,
+              headers: ProtoUtils.headers_from_proto_map(job.headers, @payload_converter) || {}
+            )
+
+            # Accept
+            add_command(
+              Bridge::Api::WorkflowCommands::WorkflowCommand.new(
+                update_response: Bridge::Api::WorkflowCommands::UpdateResponse.new(
+                  protocol_instance_id: job.protocol_instance_id,
+                  accepted: Google::Protobuf::Empty.new
+                )
+              )
+            )
+            accepted = true
+
+            # Issue update
+            result = @inbound.handle_update(input)
+
+            add_command(
+              Bridge::Api::WorkflowCommands::WorkflowCommand.new(
+                update_response: Bridge::Api::WorkflowCommands::UpdateResponse.new(
+                  protocol_instance_id: job.protocol_instance_id,
+                  completed: @payload_converter.to_payload(result)
+                )
+              )
+            )
+          rescue Exception => e # rubocop:disable Lint/RescueException
+            # Re-raise to cause task failure if this is accepted but this is not a failure exception
+            raise if accepted && !failure_exception?(e)
+
+            # Reject
+            add_command(
+              Bridge::Api::WorkflowCommands::WorkflowCommand.new(
+                update_response: Bridge::Api::WorkflowCommands::UpdateResponse.new(
+                  protocol_instance_id: job.protocol_instance_id,
+                  rejected: @failure_converter.to_failure(e, @payload_converter)
+                )
+              )
+            )
+          end
+        end
+
+        def run_workflow
+          result = @inbound.execute(
+            Temporalio::Worker::Interceptor::Workflow::ExecuteInput.new(
+              args: @workflow_arguments,
+              headers: @info.headers
+            )
+          )
+          add_command(
+            Bridge::Api::WorkflowCommands::WorkflowCommand.new(
+              complete_workflow_execution: Bridge::Api::WorkflowCommands::CompleteWorkflowExecution.new(
+                result: @payload_converter.to_payload(result)
+              )
+            )
+          )
+        end
+
+        def schedule(
+          top_level: false,
+          handler_exec: nil,
+          &
+        )
+          in_progress_handlers << handler_exec if handler_exec
+          Fiber.schedule do
+            yield
+          rescue Exception => e # rubocop:disable Lint/RescueException
+            if top_level
+              on_top_level_exception(e)
+            else
+              @current_activation_error ||= e
+            end
+          ensure
+            in_progress_handlers.delete(handler_exec) if handler_exec
+          end
+        end
+
+        def on_top_level_exception(err)
+          if err.is_a?(Workflow::ContinueAsNewError)
+            @logger.debug('Workflow requested continue as new')
+            add_command(
+              Bridge::Api::WorkflowCommands::WorkflowCommand.new(
+                continue_as_new_workflow_execution: Bridge::Api::WorkflowCommands::ContinueAsNewWorkflowExecution.new(
+                  workflow_type: if err.workflow
+                                   Workflow::Definition._workflow_type_from_workflow_parameter(err.workflow)
+                                 end,
+                  task_queue: err.task_queue,
+                  arguments: ProtoUtils.convert_to_payload_array(payload_converter, err.args),
+                  workflow_run_timeout: ProtoUtils.seconds_to_duration(err.run_timeout),
+                  workflow_task_timeout: ProtoUtils.seconds_to_duration(err.task_timeout),
+                  memo: ProtoUtils.memo_to_proto_hash(err.memo, payload_converter),
+                  headers: ProtoUtils.headers_to_proto_hash(err.headers, payload_converter),
+                  search_attributes: err.search_attributes&._to_proto,
+                  retry_policy: err.retry_policy&._to_proto
+                )
+              )
+            )
+          elsif @cancellation.canceled? && Error.canceled?(err)
+            # If cancel was ever requested and this is a cancellation or an activity/child cancellation, we add a
+            # cancel command. Technically this means that a swallowed cancel followed by, say, an activity cancel
+            # later on will show the workflow as cancelled. But this is a Temporal limitation in that cancellation is
+            # a state not an event.
+            @logger.debug('Workflow requested to cancel and properly raised cancel')
+            @logger.debug(err)
+            add_command(
+              Bridge::Api::WorkflowCommands::WorkflowCommand.new(
+                cancel_workflow_execution: Bridge::Api::WorkflowCommands::CancelWorkflowExecution.new
+              )
+            )
+          elsif failure_exception?(err)
+            @logger.debug('Workflow raised failure')
+            @logger.debug(err)
+            add_command(
+              Bridge::Api::WorkflowCommands::WorkflowCommand.new(
+                fail_workflow_execution: Bridge::Api::WorkflowCommands::FailWorkflowExecution.new(
+                  failure: @failure_converter.to_failure(err, @payload_converter)
+                )
+              )
+            )
+          else
+            @current_activation_error ||= err
+          end
+        end
+
+        def failure_exception?(err)
+          err.is_a?(Error::Failure) || err.is_a?(Timeout::Error) ||
+            @workflow_failure_exception_types&.any? { |cls| err.is_a?(cls) } ||
+            @definition_options.failure_exception_types&.any? { |cls| err.is_a?(cls) }
+        end
+
+        def with_context_frozen(&)
+          @context_frozen = true
+          yield
+        ensure
+          @context_frozen = false
+        end
+
+        def convert_handler_args(payload_array:, defn:)
+          convert_args(
+            payload_array:,
+            method_name: defn.to_invoke.is_a?(Symbol) ? defn.to_invoke : nil,
+            raw_args: defn.raw_args,
+            ignore_first_param: defn.name.nil? # Dynamic
+          )
+        end
+
+        def convert_args(payload_array:, method_name:, raw_args:, ignore_first_param: false)
+          # Just in case it is not an array
+          payload_array = payload_array.to_ary
+
+          # We want to discard extra arguments if we can. If there is a method
+          # name, try to look it up. Then, assuming there's no :rest, trim args
+          # to the amount of :req or :opt there are.
+          if method_name && @definition.workflow_class.method_defined?(method_name)
+            count = 0
+            req_count = 0
+            @definition.workflow_class.instance_method(method_name).parameters.each do |(type, _)|
+              if type == :rest
+                count = nil
+                break
+              elsif %i[req opt].include?(type)
+                count += 1
+                req_count += 1 if type == :req
+              end
+            end
+            # Fail if too few required param values, trim off excess if too many. If count is nil, it has a splat.
+            if count
+              if ignore_first_param
+                count -= 1
+                req_count -= 1
+              end
+              if req_count > payload_array.size
+                # We have to fail here instead of let Ruby fail the invocation because some handlers, such as signals,
+                # want to log and ignore invalid arguments instead of fail and if we used Ruby failure, we can't
+                # differentiate between too-few-param caused by us or somewhere else by a user.
+                raise ArgumentError, "wrong number of required arguments for #{method_name} " \
+                                     "(given #{payload_array.size}, expected #{req_count})"
+              end
+              payload_array = payload_array.take(count)
+            end
+          end
+
+          # Convert
+          if raw_args
+            payload_array.map { |p| Converters::RawValue.new(p) }
+          else
+            ProtoUtils.convert_from_payload_array(@payload_converter, payload_array)
+          end
+        end
+
+        def workflow_metadata
+          Temporalio::Api::Sdk::V1::WorkflowMetadata.new(
+            definition: Temporalio::Api::Sdk::V1::WorkflowDefinition.new(
+              type: info.workflow_type,
+              query_definitions: query_handlers.values.map do |defn|
+                Temporalio::Api::Sdk::V1::WorkflowInteractionDefinition.new(
+                  name: defn.name || '', description: defn.description || ''
+                )
+              end,
+              signal_definitions: signal_handlers.values.map do |defn|
+                Temporalio::Api::Sdk::V1::WorkflowInteractionDefinition.new(
+                  name: defn.name || '', description: defn.description || ''
+                )
+              end,
+              update_definitions: update_handlers.values.map do |defn|
+                Temporalio::Api::Sdk::V1::WorkflowInteractionDefinition.new(
+                  name: defn.name || '', description: defn.description || ''
+                )
+              end
+            ),
+            current_details: current_details || ''
+          )
+        end
+
+        def scoped_logger_info
+          @scoped_logger_info ||= {
+            attempt: info.attempt,
+            namespace: info.namespace,
+            run_id: info.run_id,
+            task_queue: info.task_queue,
+            workflow_id: info.workflow_id,
+            workflow_type: info.workflow_type
+          }
+          # Append update info if there is any
+          update_info = Fiber[:__temporal_update_info]
+          return @scoped_logger_info unless update_info
+
+          @scoped_logger_info.merge({ update_id: update_info.id, update_name: update_info.name })
+        end
+
+        def warn_on_any_unfinished_handlers
+          updates, signals = in_progress_handlers.select do |h|
+            h.unfinished_policy == Workflow::HandlerUnfinishedPolicy::WARN_AND_ABANDON
+          end.partition(&:update_id)
+
+          unless updates.empty?
+            updates_str = JSON.generate(updates.map { |u| { name: u.name, id: u.update_id } })
+            warn(
+              "[TMPRL1102] Workflow #{info.workflow_id} finished while update handlers are still running. This may " \
+              'have interrupted work that the update handler was doing, and the client that sent the update will ' \
+              "receive a 'workflow execution already completed' RPCError instead of the update result. You can wait " \
+              'for all update and signal handlers to complete by using ' \
+              '`Temporalio::Workflow.wait_condition { Temporalio::Workflow.handlers_finished? }`. ' \
+              'Alternatively, if both you and the clients sending the update are okay with  interrupting running ' \
+              'handlers when the workflow finishes, and causing clients to receive errors, then you can disable this ' \
+              'warning via the update handler definition: ' \
+              '`workflow_update unfinished_policy: Temporalio::Workflow::HandlerUnfinishedPolicy.ABANDON`. ' \
+              "The following updates were unfinished (and warnings were not disabled for their handler): #{updates_str}"
+            )
+          end
+
+          return if signals.empty?
+
+          signals_str = JSON.generate(signals.group_by(&:name)
+            .transform_values(&:size).sort_by { |_, v| -v }.map { |name, count| { name:, count: } })
+          warn(
+            "[TMPRL1102] Workflow #{info.workflow_id} finished while signal handlers are still running. This may " \
+            'have interrupted work that the signal handler was doing. You can wait for all update and signal ' \
+            'handlers to complete by using ' \
+            '`Temporalio::Workflow.wait_condition { Temporalio::Workflow.handlers_finished? }`. ' \
+            'Alternatively, if both you and the clients sending the signal are okay with  interrupting running ' \
+            'handlers when the workflow finishes, then you can disable this warning via the signal handler ' \
+            'definition: ' \
+            '`workflow_signal unfinished_policy: Temporalio::Workflow::HandlerUnfinishedPolicy.ABANDON`. ' \
+            "The following signals were unfinished (and warnings were not disabled for their  handler): #{signals_str}"
+          )
+        end
+      end
+    end
+  end
+end