Show HN: Update on Dispytch: Just Got Dynamic Topics – Event Handling Leveled Up

1 comments

• e1-m 7 hours ago
  Hey folks, quick update! I just shipped a new version of [*Dispytch*](https://github.com/e1-m/dispytch) — async Python framework for building event-driven services.

  # What Dispytch Does

  Dispytch makes it easy to build services that react to events — whether they're coming from Kafka, RabbitMQ, Redis or some other broker. You define event types as Pydantic models and wire up handlers with dependency injection. Dispytch handles validation, retries, and routing out of the box, so you can focus on the logic.

  # Comparison

  |Framework|Focus|Notes| |:-|:-|:-| |Celery|Task queues|Great for backgroud processing| |Faust|Kafka streams|Powerful, but streaming-centric| |Nameko|RPC services|Sync-first, heavy| |FastAPI|HTTP APIs|Not for event processing| |FastStream|Stream pipelines|Built around streams—great for data pipelines.| |*Dispytch*|Event handling|Event-centric and reactive, designed for clear event-driven services.|

  # Quick API Example

  # Handler

      user_events.handler(topic='user_events', event='user_registered')
      async def handle_user_registered(
              event: Event[UserCreatedEvent],
              user_service: Annotated[UserService, Dependency(get_user_service)]
      ):
          user = event.body.user
          timestamp = event.body.timestamp
      
          print(f"[User Registered] {user.id} - {user.email} at {timestamp}")
      
          await user_service.do_smth_with_the_user(event.body.user)
  

  # Emitter

      async def example_emit(emitter):
         await emitter.emit(
             UserRegistered(
                 user=User(
                     id=str(uuid.uuid4()),
                     email="example@mail.com",
                     name="John Doe",
                 ),
                 timestamp=int(datetime.now().timestamp()),
             )
         )
  

  # What’s New?

  *Redis Pub/Sub support* You can now plug Redis into Dispytch and start consuming events without spinning up Kafka or RabbitMQ. Perfect for lightweight setups.

  *Dynamic Topics* Handlers can now use topic segments as function arguments — e.g., match `"user.{user_id}.notification"` and get `user_id` injected automatically. Clean and type-safe thanks to Pydantic validation.

  Try it out:

      uv add dispytch
  
   Docs and examples in the repo: [https://github.com/e1-m/dispytch](https://github.com/e1-m/dispytch)
  

  Feedback, bug reports, feature requests — all welcome. Still early, still evolving

  Thanks for checking it out!