Python asyncio/discord.py - 循环退出，任务被破坏，但它处于待处理状态

Posted 2023-03-14

【中文标题】Python asyncio/discord.py - 循环退出，任务被破坏，但它处于待处理状态

【英文标题】：Python asyncio/discord.py - Loop exits with Task was destroyed but it is pending

【发布时间】：2017-12-17 08:36:29

【问题描述】：

与我的问题相关的代码：

discordbot.py

import discord, main, websockets
from discord.ext import commands

TOKEN = ''

description = '''Description'''
bot = commands.Bot(command_prefix='!', description=description)

@bot.event
async def on_ready():
    print('----------------------------')
    print('Connected!')
    print('Logged in as: 0'.format(bot.user.name))
    print('Bot ID: 0'.format(bot.user.id))
    print('----------------------------')


@bot.command()
async def wall(coin, desired_multiplier):
    try:
        desired_multiplier = float(desired_multiplier)
    except:
        await bot.say("That's not a percent!")
    desired_multiplier = float("0:.1f".format(desired_multiplier))
    try:
        if desired_multiplier <= 2.0:
            #await bot.say('Calculating...')
            volume, rate, last_price = main.get_sells(coin, desired_multiplier)
            total_btc = sum(volume)
            total_btc = float("0:.3f".format(total_btc))
            await bot.say('Success')
            print('Total of 0 BTC to reach a 1x multiplier for 2'.format(total_btc, desired_multiplier, coin.upper()))
        else:
            await bot.say('Please use a multiplier under 2x, apparently I can\'t handle more than that.')
    except:
        await bot.say('Error: Please make sure the coin is registered on Bittrex!')

bot.run(TOKEN)

ma​​in.py

from bittrex import Bittrex
import requests

bittrex = Bittrex('', '')
BUY_ORDERBOOK = 'buy'
SELL_ORDERBOOK = 'sell'
BOTH_ORDERBOOK = 'both'

def total_btc(quantity, rate):
    total = rate * quantity
    return total

def get_sells(coin, desired_multiplier):
    volume = []
    sells = bittrex.get_orderbook('BTC-0'.format(coin), SELL_ORDERBOOK)
    if sells['success'] is False:
        print(sells['message'], '#########################################################')
    for sell in range(10000):
        order = sells['result'][sell]
        quantity = order['Quantity']
        rate = order['Rate']
        total = total_btc(quantity, rate)
        volume.append(total)
        last_price = bittrex.get_ticker('BTC-0'.format(coin))['result']['Last']
        multiplier = rate / last_price
        multiplier = float("0:.1f".format(multiplier))
        print('0 | 1'.format(sell, multiplier))
        if multiplier != desired_multiplier:
            continue
        else:
            return volume, rate, last_price

错误和回溯：

Task was destroyed but it is pending!
task: <Task pending coro=<Client._run_event() running at C:\Users\logicmn\D
ocuments\discordenv\lib\site-packages\discord\client.py:307> wait_for=<Future pe
nding cb=[BaseSelectorEventLoop._sock_connect_done(704)(), <TaskWakeupMethWrappe
r object at 0x0000000003CE2258>()]>>
Exception ignored in: <generator object Bot.on_message at 0x0000000003CDC048>
Traceback (most recent call last):
  File "C:\Users\logicmn\Documents\discordenv\lib\site-packages\discord\ext
\commands\bot.py", line 857, in on_message
    yield from self.process_commands(message)
  File "C:\Users\logicmn\Documents\discordenv\lib\site-packages\discord\ext
\commands\bot.py", line 848, in process_commands
    ctx.command.dispatch_error(e, ctx)
  File "C:\Users\logicmn\Documents\discordenv\lib\site-packages\discord\ext
\commands\core.py", line 164, in dispatch_error
    ctx.bot.dispatch('command_error', error, ctx)
  File "C:\Users\logicmn\Documents\discordenv\lib\site-packages\discord\ext
\commands\bot.py", line 262, in dispatch
    super().dispatch(event_name, *args, **kwargs)
  File "C:\Users\logicmn\Documents\discordenv\lib\site-packages\discord\cli
ent.py", line 325, in dispatch
    compat.create_task(self._run_event(method, *args, **kwargs), loop=self.loop)

  File "c:\users\logicmn\appdata\local\continuum\anaconda3\Lib\asyncio\task
s.py", line 512, in ensure_future
    task = loop.create_task(coro_or_future)
  File "c:\users\logicmn\appdata\local\continuum\anaconda3\Lib\asyncio\base
_events.py", line 282, in create_task
    self._check_closed()
  File "c:\users\logicmn\appdata\local\continuum\anaconda3\Lib\asyncio\base
_events.py", line 357, in _check_closed
    raise RuntimeError('Event loop is closed')
RuntimeError: Event loop is closed

This seems relevant 但我尝试了该解决方案但无法使其正常工作。我确实相信这个问题是因为不和谐客户端模块需要每 60 秒控制一次。由于我的函数 get_sells() 需要大约 70 秒，因此 asyncio 会中断。

【参考方案1】：

discord 客户端显然需要每分钟左右进行一次控制，而您的 get_sells 函数的阻塞不止于此。

这是因为您将同步 bittrex 客户端与异步库 (discord) 混合在一起。您需要对两者使用相同的范例：discord 客户端基于 asyncio，因此您需要为 bittrex 找到一个基于 asyncio 的库（或编写一个）。

这个新库应该公开一些与您正在使用的函数类似的协程。所以你会调用sells = await bittrex_aio.get_orderbook(…)，而不是sells = bittrex.get_orderbook(…)，def get_sells(…) 会变成协程async def get_sells(…)。

这样，您将能够将您的机器人更改为使用await main.get_sells，这意味着不和谐客户端将能够控制您正在等待get_sells 的结果。

我不是故意在这里详细介绍 asyncio，但阅读它可能会有所帮助，例如：https://docs.python.org/3/library/asyncio.html