pynenc.runner.thread_runner

Module Contents

Classes

ThreadInfo
ThreadRunner	ThreadRunner is a concrete implementation of BaseRunner that executes tasks in separate threads.

API

class pynenc.runner.thread_runner.ThreadInfo

Bases: typing.NamedTuple

thread: threading.Thread

None

invocation: pynenc.invocation.dist_invocation.DistributedInvocation

None

class pynenc.runner.thread_runner.ThreadRunner(app: pynenc.app.Pynenc, runner_cache: dict | None = None, runner_context: pynenc.runner.runner_context.RunnerContext | None = None)

Bases: pynenc.runner.base_runner.BaseRunner

ThreadRunner is a concrete implementation of BaseRunner that executes tasks in separate threads.

It manages task invocations, handling their execution and lifecycle within individual threads. This runner is suitable for I/O-bound tasks and scenarios where shared memory between tasks is required.

Initialization

conf() → pynenc.conf.config_runner.ConfigThreadRunner

static mem_compatible() → bool

Indicates if the runner is compatible with in-memory components.

Returns:

True, as each task is executed in a separate thread with shared memory.

property max_parallel_slots: int

The maximum number of parallel tasks that the runner can handle.

Returns:

An integer representing the maximum number of parallel tasks.

get_active_child_runner_ids() → list[str]

ThreadRunner doesn’t spawn child processes, so returns empty list.

_log_shutdown(signum: int | None) → None

_on_start() → None

Internal method called when the ThreadRunner starts. Initializes the data infrastructures for managing invocations and threads.

_on_stop() → None

Internal method called when the ThreadRunner stops. Joins all running threads and updates their invocation statuses.

For alive threads: reroute first while we still own the RUNNING status, then join. If we joined first the task would complete as SUCCESS and the RUNNING→KILLED transition would no longer be valid. For already-dead threads: join (no-op) then reroute, which silently ignores the final status.

_on_stop_runner_loop() → None

Internal method called after receiving a signal to stop the runner loop.

_reclaim_available_slots() → int

Joins finished threads and returns the number of available thread slots.

Returns:

An integer representing available thread slots.

runner_loop_iteration() → None

Executes one iteration of the ThreadRunner loop. Handles the execution and monitoring of task invocations in separate threads.

_waiting_for_results(running_invocation_id: pynenc.identifiers.invocation_id.InvocationId, result_invocation_ids: list[pynenc.identifiers.invocation_id.InvocationId], runner_args: dict[str, Any] | None = None) → None

Handles invocations waiting for results by polling a local final cache instead of pausing threads.

Parameters:

• running_invocation_id – The invocation that is waiting for results.

• result_invocation_ids – A list of invocations whose results are being awaited.

• runner_args – Additional arguments required for the ThreadRunner.