Source code for aiogram.dispatcher.dispatcher

from __future__ import annotations

import asyncio
import contextvars
import signal
import warnings
from asyncio import CancelledError, Event, Future, Lock
from contextlib import suppress
from typing import Any, AsyncGenerator, Dict, List, Optional, Set, Union

from .. import loggers
from import Bot
from ..exceptions import TelegramAPIError
from ..fsm.middleware import FSMContextMiddleware
from import BaseEventIsolation, BaseStorage
from import DisabledEventIsolation, MemoryStorage
from ..fsm.strategy import FSMStrategy
from ..methods import GetUpdates, TelegramMethod
from ..methods.base import TelegramType
from ..types import Update, User
from ..types.base import UNSET, UNSET_TYPE
from ..types.update import UpdateTypeLookupError
from ..utils.backoff import Backoff, BackoffConfig
from .event.bases import UNHANDLED, SkipHandler
from .event.telegram import TelegramEventObserver
from .middlewares.error import ErrorsMiddleware
from .middlewares.user_context import UserContextMiddleware
from .router import Router

DEFAULT_BACKOFF_CONFIG = BackoffConfig(min_delay=1.0, max_delay=5.0, factor=1.3, jitter=0.1)

[docs] class Dispatcher(Router): """ Root router """
[docs] def __init__( self, *, # * - Preventing to pass instance of Bot to the FSM storage storage: Optional[BaseStorage] = None, fsm_strategy: FSMStrategy = FSMStrategy.USER_IN_CHAT, events_isolation: Optional[BaseEventIsolation] = None, disable_fsm: bool = False, name: Optional[str] = None, **kwargs: Any, ) -> None: """ Root router :param storage: Storage for FSM :param fsm_strategy: FSM strategy :param events_isolation: Events isolation :param disable_fsm: Disable FSM, note that if you disable FSM then you should not use storage and events isolation :param kwargs: Other arguments, will be passed as keyword arguments to handlers """ super(Dispatcher, self).__init__(name=name) if storage and not isinstance(storage, BaseStorage): raise TypeError( f"FSM storage should be instance of 'BaseStorage' not {type(storage).__name__}" ) # Telegram API provides originally only one event type - Update # For making easily interactions with events here is registered handler which helps # to separate Update to different event types like Message, CallbackQuery etc. self.update = self.observers["update"] = TelegramEventObserver( router=self, event_name="update" ) self.update.register(self._listen_update) # Error handlers should work is out of all other functions # and should be registered before all others middlewares self.update.outer_middleware(ErrorsMiddleware(self)) # User context middleware makes small optimization for all other builtin # middlewares via caching the user and chat instances in the event context self.update.outer_middleware(UserContextMiddleware()) # FSM middleware should always be registered after User context middleware # because here is used context from previous step self.fsm = FSMContextMiddleware( storage=storage or MemoryStorage(), strategy=fsm_strategy, events_isolation=events_isolation or DisabledEventIsolation(), ) if not disable_fsm: # Note that when FSM middleware is disabled, the event isolation is also disabled # Because the isolation mechanism is a part of the FSM self.update.outer_middleware(self.fsm) self.shutdown.register(self.fsm.close) self.workflow_data: Dict[str, Any] = kwargs self._running_lock = Lock() self._stop_signal: Optional[Event] = None self._stopped_signal: Optional[Event] = None self._handle_update_tasks: Set[asyncio.Task[Any]] = set()
def __getitem__(self, item: str) -> Any: return self.workflow_data[item] def __setitem__(self, key: str, value: Any) -> None: self.workflow_data[key] = value def __delitem__(self, key: str) -> None: del self.workflow_data[key] def get(self, key: str, /, default: Optional[Any] = None) -> Optional[Any]: return self.workflow_data.get(key, default) @property def storage(self) -> BaseStorage: return @property def parent_router(self) -> Optional[Router]: """ Dispatcher has no parent router and can't be included to any other routers or dispatchers :return: """ return None # noqa: RET501 @parent_router.setter def parent_router(self, value: Router) -> None: """ Dispatcher is root Router then configuring parent router is not allowed :param value: :return: """ raise RuntimeError("Dispatcher can not be attached to another Router.")
[docs] async def feed_update(self, bot: Bot, update: Update, **kwargs: Any) -> Any: """ Main entry point for incoming updates Response of this method can be used as Webhook response :param bot: :param update: """ loop = asyncio.get_running_loop() handled = False start_time = loop.time() if != bot: # Re-mounting update to the current bot instance for making possible to # use it in shortcuts. # Here is update is re-created because we need to propagate context to # all nested objects and attributes of the Update, but it # is impossible without roundtrip to JSON :( # The preferred way is that pass already mounted Bot instance to this update # before call feed_update method update = Update.model_validate(update.model_dump(), context={"bot": bot}) try: response = await self.update.wrap_outer_middleware( self.update.trigger, update, { **self.workflow_data, **kwargs, "bot": bot, }, ) handled = response is not UNHANDLED return response finally: finish_time = loop.time() duration = (finish_time - start_time) * 1000 "Update id=%s is %s. Duration %d ms by bot id=%d", update.update_id, "handled" if handled else "not handled", duration,, )
[docs] async def feed_raw_update(self, bot: Bot, update: Dict[str, Any], **kwargs: Any) -> Any: """ Main entry point for incoming updates with automatic Dict->Update serializer :param bot: :param update: :param kwargs: """ parsed_update = Update.model_validate(update, context={"bot": bot}) return await self.feed_update(bot=bot, update=parsed_update, **kwargs)
@classmethod async def _listen_updates( cls, bot: Bot, polling_timeout: int = 30, backoff_config: BackoffConfig = DEFAULT_BACKOFF_CONFIG, allowed_updates: Optional[List[str]] = None, ) -> AsyncGenerator[Update, None]: """ Endless updates reader with correctly handling any server-side or connection errors. So you may not worry that the polling will stop working. """ backoff = Backoff(config=backoff_config) get_updates = GetUpdates(timeout=polling_timeout, allowed_updates=allowed_updates) kwargs = {} if bot.session.timeout: # Request timeout can be lower than session timeout and that's OK. # To prevent false-positive TimeoutError we should wait longer than polling timeout kwargs["request_timeout"] = int(bot.session.timeout + polling_timeout) failed = False while True: try: updates = await bot(get_updates, **kwargs) except Exception as e: failed = True # In cases when Telegram Bot API was inaccessible don't need to stop polling # process because some developers can't make auto-restarting of the script loggers.dispatcher.error("Failed to fetch updates - %s: %s", type(e).__name__, e) # And also backoff timeout is best practice to retry any network activity loggers.dispatcher.warning( "Sleep for %f seconds and try again... (tryings = %d, bot id = %d)", backoff.next_delay, backoff.counter,, ) await backoff.asleep() continue # In case when network connection was fixed let's reset the backoff # to initial value and then process updates if failed: "Connection established (tryings = %d, bot id = %d)", backoff.counter,, ) backoff.reset() failed = False for update in updates: yield update # The getUpdates method returns the earliest 100 unconfirmed updates. # To confirm an update, use the offset parameter when calling getUpdates # All updates with update_id less than or equal to offset will be marked # as confirmed on the server and will no longer be returned. get_updates.offset = update.update_id + 1 async def _listen_update(self, update: Update, **kwargs: Any) -> Any: """ Main updates listener Workflow: - Detect content type and propagate to observers in current router - If no one filter is pass - propagate update to child routers as Update :param update: :param kwargs: :return: """ try: update_type = update.event_type event = update.event except UpdateTypeLookupError as e: warnings.warn( "Detected unknown update type.\n" "Seems like Telegram Bot API was updated and you have " "installed not latest version of aiogram framework" f"\nUpdate: {update.model_dump_json(exclude_unset=True)}", RuntimeWarning, ) raise SkipHandler() from e kwargs.update(event_update=update) return await self.propagate_event(update_type=update_type, event=event, **kwargs) @classmethod async def silent_call_request(cls, bot: Bot, result: TelegramMethod[Any]) -> None: """ Simulate answer into WebHook :param bot: :param result: :return: """ try: await bot(result) except TelegramAPIError as e: # In due to WebHook mechanism doesn't allow getting response for # requests called in answer to WebHook request. # Need to skip unsuccessful responses. # For debugging here is added logging. loggers.event.error("Failed to make answer: %s: %s", e.__class__.__name__, e) async def _process_update( self, bot: Bot, update: Update, call_answer: bool = True, **kwargs: Any ) -> bool: """ Propagate update to event listeners :param bot: instance of Bot :param update: instance of Update :param call_answer: need to execute response as Telegram method (like answer into webhook) :param kwargs: contextual data for middlewares, filters and handlers :return: status """ try: response = await self.feed_update(bot, update, **kwargs) if call_answer and isinstance(response, TelegramMethod): await self.silent_call_request(bot=bot, result=response) return response is not UNHANDLED except Exception as e: loggers.event.exception( "Cause exception while process update id=%d by bot id=%d\n%s: %s", update.update_id,, e.__class__.__name__, e, ) return True # because update was processed but unsuccessful async def _polling( self, bot: Bot, polling_timeout: int = 30, handle_as_tasks: bool = True, backoff_config: BackoffConfig = DEFAULT_BACKOFF_CONFIG, allowed_updates: Optional[List[str]] = None, **kwargs: Any, ) -> None: """ Internal polling process :param bot: :param kwargs: :return: """ user: User = await "Run polling for bot @%s id=%d - %r", user.username,, user.full_name ) try: async for update in self._listen_updates( bot, polling_timeout=polling_timeout, backoff_config=backoff_config, allowed_updates=allowed_updates, ): handle_update = self._process_update(bot=bot, update=update, **kwargs) if handle_as_tasks: handle_update_task = asyncio.create_task(handle_update) self._handle_update_tasks.add(handle_update_task) handle_update_task.add_done_callback(self._handle_update_tasks.discard) else: await handle_update finally: "Polling stopped for bot @%s id=%d - %r", user.username,, user.full_name ) async def _feed_webhook_update(self, bot: Bot, update: Update, **kwargs: Any) -> Any: """ The same with `Dispatcher.process_update()` but returns real response instead of bool """ try: return await self.feed_update(bot, update, **kwargs) except Exception as e: loggers.event.exception( "Cause exception while process update id=%d by bot id=%d\n%s: %s", update.update_id,, e.__class__.__name__, e, ) raise async def feed_webhook_update( self, bot: Bot, update: Union[Update, Dict[str, Any]], _timeout: float = 55, **kwargs: Any ) -> Optional[TelegramMethod[TelegramType]]: if not isinstance(update, Update): # Allow to use raw updates update = Update.model_validate(update, context={"bot": bot}) ctx = contextvars.copy_context() loop = asyncio.get_running_loop() waiter = loop.create_future() def release_waiter(*_: Any) -> None: if not waiter.done(): waiter.set_result(None) timeout_handle = loop.call_later(_timeout, release_waiter) process_updates: Future[Any] = asyncio.ensure_future( self._feed_webhook_update(bot=bot, update=update, **kwargs) ) process_updates.add_done_callback(release_waiter, context=ctx) def process_response(task: Future[Any]) -> None: warnings.warn( "Detected slow response into webhook.\n" "Telegram is waiting for response only first 60 seconds and then re-send update.\n" "For preventing this situation response into webhook returned immediately " "and handler is moved to background and still processing update.", RuntimeWarning, ) try: result = task.result() except Exception as e: raise e if isinstance(result, TelegramMethod): asyncio.ensure_future(self.silent_call_request(bot=bot, result=result)) try: try: await waiter except CancelledError: # pragma: no cover process_updates.remove_done_callback(release_waiter) process_updates.cancel() raise if process_updates.done(): # TODO: handle exceptions response: Any = process_updates.result() if isinstance(response, TelegramMethod): return response else: process_updates.remove_done_callback(release_waiter) process_updates.add_done_callback(process_response, context=ctx) finally: timeout_handle.cancel() return None
[docs] async def stop_polling(self) -> None: """ Execute this method if you want to stop polling programmatically :return: """ if not self._running_lock.locked(): raise RuntimeError("Polling is not started") if not self._stop_signal or not self._stopped_signal: return self._stop_signal.set() await self._stopped_signal.wait()
def _signal_stop_polling(self, sig: signal.Signals) -> None: if not self._running_lock.locked(): return loggers.dispatcher.warning("Received %s signal", if not self._stop_signal: return self._stop_signal.set()
[docs] async def start_polling( self, *bots: Bot, polling_timeout: int = 10, handle_as_tasks: bool = True, backoff_config: BackoffConfig = DEFAULT_BACKOFF_CONFIG, allowed_updates: Optional[Union[List[str], UNSET_TYPE]] = UNSET, handle_signals: bool = True, close_bot_session: bool = True, **kwargs: Any, ) -> None: """ Polling runner :param bots: Bot instances (one or more) :param polling_timeout: Long-polling wait time :param handle_as_tasks: Run task for each event and no wait result :param backoff_config: backoff-retry config :param allowed_updates: List of the update types you want your bot to receive By default, all used update types are enabled (resolved from handlers) :param handle_signals: handle signals (SIGINT/SIGTERM) :param close_bot_session: close bot sessions on shutdown :param kwargs: contextual data :return: """ if not bots: raise ValueError("At least one bot instance is required to start polling") if "bot" in kwargs: raise ValueError( "Keyword argument 'bot' is not acceptable, " "the bot instance should be passed as positional argument" ) async with self._running_lock: # Prevent to run this method twice at a once if self._stop_signal is None: self._stop_signal = Event() if self._stopped_signal is None: self._stopped_signal = Event() if allowed_updates is UNSET: allowed_updates = self.resolve_used_update_types() self._stop_signal.clear() self._stopped_signal.clear() if handle_signals: loop = asyncio.get_running_loop() with suppress(NotImplementedError): # pragma: no cover # Signals handling is not supported on Windows # It also can't be covered on Windows loop.add_signal_handler( signal.SIGTERM, self._signal_stop_polling, signal.SIGTERM ) loop.add_signal_handler( signal.SIGINT, self._signal_stop_polling, signal.SIGINT ) workflow_data = { "dispatcher": self, "bots": bots, **self.workflow_data, **kwargs, } if "bot" in workflow_data: workflow_data.pop("bot") await self.emit_startup(bot=bots[-1], **workflow_data)"Start polling") try: tasks: List[asyncio.Task[Any]] = [ asyncio.create_task( self._polling( bot=bot, handle_as_tasks=handle_as_tasks, polling_timeout=polling_timeout, backoff_config=backoff_config, allowed_updates=allowed_updates, **workflow_data, ) ) for bot in bots ] tasks.append(asyncio.create_task(self._stop_signal.wait())) done, pending = await asyncio.wait(tasks, return_when=asyncio.FIRST_COMPLETED) for task in pending: # (mostly) Graceful shutdown unfinished tasks task.cancel() with suppress(CancelledError): await task # Wait finished tasks to propagate unhandled exceptions await asyncio.gather(*done) finally:"Polling stopped") try: await self.emit_shutdown(bot=bots[-1], **workflow_data) finally: if close_bot_session: await asyncio.gather(*(bot.session.close() for bot in bots)) self._stopped_signal.set()
[docs] def run_polling( self, *bots: Bot, polling_timeout: int = 10, handle_as_tasks: bool = True, backoff_config: BackoffConfig = DEFAULT_BACKOFF_CONFIG, allowed_updates: Optional[Union[List[str], UNSET_TYPE]] = UNSET, handle_signals: bool = True, close_bot_session: bool = True, **kwargs: Any, ) -> None: """ Run many bots with polling :param bots: Bot instances (one or more) :param polling_timeout: Long-polling wait time :param handle_as_tasks: Run task for each event and no wait result :param backoff_config: backoff-retry config :param allowed_updates: List of the update types you want your bot to receive :param handle_signals: handle signals (SIGINT/SIGTERM) :param close_bot_session: close bot sessions on shutdown :param kwargs: contextual data :return: """ with suppress(KeyboardInterrupt): return self.start_polling( *bots, **kwargs, polling_timeout=polling_timeout, handle_as_tasks=handle_as_tasks, backoff_config=backoff_config, allowed_updates=allowed_updates, handle_signals=handle_signals, close_bot_session=close_bot_session, ) )