我正在尝试使用asyncio编写HTTP / 1服务器,我正在尝试让它处理HTTP keep-alives。我有以下一点代码。
import re
import socket
import asyncio
async def request_handler(reader, writer):
try:
keep_alive = True
while keep_alive:
keep_alive = False
while True:
print('Awaiting data')
line = await reader.readline()
print('Finished await got %s' % line)
if not line.rstrip(b'\r\n'):
break
if re.match(rb'connection:\s*keep-alive', line, re.I):
keep_alive = True
writer.write(b'HTTP/1.1 200 OK\r\n\r\n<h1>My web page</h1>\r\n')
await writer.drain()
finally:
writer.close()
if __name__ == '__main__':
loop = asyncio.get_event_loop()
coro = asyncio.start_server(request_handler, '', 8888, family=socket.AF_UNSPEC, loop=loop, limit=2048)
server = loop.run_until_complete(coro)
try:
loop.run_forever()
finally:
server.close()
loop.run_until_complete(server.wait_closed())
loop.close()
对于curl等工具的请求,请求可以在服务器挂起的情况下正确提供。
但是,尝试在浏览器中加载URL会导致服务器永远不会终止连接。浏览器尝试请求两个资源,一个来自/
,另一个来自/favicon.ico
,并且HTTP keep-alives用于请求。 (可以通过开发人员工具查看此信息。)
我尝试打印服务器收到的数据。但是,服务器似乎永远不会收到第二个请求的数据:
Awaiting data
Finished await got b'GET / HTTP/1.1\r\n'
Awaiting data
Finished await got b'Host: localhost:8888\r\n'
Awaiting data
Finished await got b'Connection: keep-alive\r\n'
Awaiting data
Finished await got b'Upgrade-Insecure-Requests: 1\r\n'
Awaiting data
Finished await got b'User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10_13_6) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/68.0.3440.106 Safari/537.36\r\n'
Awaiting data
Finished await got b'Accept: text/html,application/xhtml+xml,application/xml;q=0.9,image/webp,image/apng,*/*;q=0.8\r\n'
Awaiting data
Finished await got b'Accept-Encoding: gzip, deflate, br\r\n'
Awaiting data
Finished await got b'Accept-Language: en-US,en;q=0.9\r\n'
Awaiting data
Finished await got b'\r\n'
Awaiting data
谁能告诉我这是什么问题?
使用Keep-Alive时,响应必须包含Content-Length
标头(或使用更复杂的chunked传输编码)。没有它,客户端别无选择,只能等待连接关闭 - 这绝不会因为keep_alive
设置为true而发生。例如,如果您更改这样的编写代码:
body = b'<h1>My web page</h1>\r\n'
writer.write(b'HTTP/1.1 200 OK\r\nContent-Length: %d\r\n\r\n' % len(body))
writer.write(body)
......连接不再悬挂。
curl
的工作原理是因为它没有指定Keep-Alive
(因为在命令行中只指定了一个URL),因此您的代码会关闭连接并且不需要内容长度。