Core message routing layer for Jupyter AI

[3coins]

This extension provides the foundational message routing functionality for Jupyter AI. It automatically detects new chat sessions and routes messages to registered callbacks based on message type (slash commands vs regular messages). Extensions can register callbacks to handle specific chat events without needing to manage chat lifecycle directly.

Usage

Basic MessageRouter Setup

# The router is available in other extensions via settings
router = self.serverapp.web_app.settings.get("jupyter-ai", {}).get("router")

# Register callbacks for different event types
def on_new_chat(room_id: str, ychat: YChat):
    print(f"New chat connected: {room_id}")

def on_slash_command(room_id: str, message: Message):
    print(f"Slash command in {room_id}: {message.body}")

def on_regular_message(room_id: str, message: Message):`
    print(f"Regular message in {room_id}: {message.body}")

# Register the callbacks
router.observe_chat_init(on_new_chat)
router.observe_slash_cmd_msg("room-id", on_slash_command)
router.observe_chat_msg("room-id", on_regular_message)

Message Flow

1. Router detects new chats - Automatically listens for chat room initialization events
2. Router connects chats - Establishes observers on YChat message streams
3. Router routes messages - Calls appropriate callbacks based on message type (slash vs regular)
4. Extensions respond - Your callbacks receive room_id and message data

Available Methods

• observe_chat_init(callback) - Called when new chat sessions are initialized with (room_id, ychat)
• observe_slash_cmd_msg(room_id, callback) - Called for messages starting with / in a specific room
• observe_chat_msg(room_id, callback) - Called for regular (non-slash) messages in a specific room