今天拉了一个使用了tornado的项目在本地跑，按照源码作者的步骤配置完，运行，直接报错了，要求环境Python3.6+，我装的是Python3.8，理论上应该直接正常运行的，报错信息：

Traceback (most recent call last):
  File "ice_server.py", line 150, in
    RunServer.run_server(port=p, host=h)
  File "ice_server.py", line 125, in run_server
    tornado_server.start()
  File "D:\PycharmProjects\ice\venv\lib\site-packages\tornado\tcpserver.py", line 244, in start
    self.add_sockets(sockets)
  File "D:\PycharmProjects\ice\venv\lib\site-packages\tornado\tcpserver.py", line 165, in add_sockets
    self._handlers[sock.fileno()] = add_accept_handler(
  File "D:\PycharmProjects\ice\venv\lib\site-packages\tornado\netutil.py", line 279, in add_accept_handler
    io_loop.add_handler(sock, accept_handler, IOLoop.READ)
  File "D:\PycharmProjects\ice\venv\lib\site-packages\tornado\platform\asyncio.py", line 100, in add_handler
    self.asyncio_loop.add_reader(fd, self._handle_events, fd, IOLoop.READ)
  File "C:\Users\huan\AppData\Local\Programs\Python\Python38\lib\asyncio\events.py", line 501, in add_reader
    raise NotImplementedError
NotImplementedError

一番谷歌原来对于这个问题tornado的参与者们已经收到了很多反馈，有个回复里这么说：

Python 3.8 asyncio is going to make the "proactor" event loop the default, instead of the current "selector" event loop. This is a problem for Tornado because the proactor event loop doesn't support the unix-style add_reader APIs that Tornado uses.

Anyone using Tornado 5+ on windows with python 3.8 will need to configure asyncio to use the selector event loop; we'll have to document this. We should also try to detect the use of a proactor event loop and give a clear error message

大概意思Python3.8的asyncio改变了循环方式，因为这种方式在windows上不支持相应的add_reader APIs，就会抛出NotImplementedError错误。

解决办法

找到这个项目使用的python环境的lib\site-packages，做下面的修改，在path-to-python\lib\site-packages\tornado\platform\asyncio.py开头添加代码：

import sys

if sys.platform == 'win32':
  asyncio.set_event_loop_policy(asyncio.WindowsSelectorEventLoopPolicy())

这样就可以正常运行了。

总结

到此这篇关于Python3.8运行tornado项目报NotImplementedError错误的文章就介绍到这了,更多相关Python3.8运行tornado项目报错内容请搜索python博客以前的文章或继续浏览下面的相关文章希望大家以后多多支持python博客！