一、

首先创建三个服务器，代码如下：

flask服务器.py

 1 from flask import Flask
 2 import time
 3 
 4 app = Flask(__name__)
 5 
 6 @app.route('/bobo')
 7 def index_bobo():
 8     time.sleep(2)
 9     return 'Hello bobo'
10 
11 @app.route('/jay')
12 def index_jay():
13     time.sleep(2)
14     return 'Hello jay'
15 
16 @app.route('/tom')
17 def index_tom():
18     time.sleep(2)
19     return 'Hello tom'
20 
21 if __name__=="__main__":
22     app.run(threaded=True)

运行：

 

 然后开始编写多任务协程

 1 import requests
 2 import asyncio
 3 import time
 4 
 5 
 6 start=time.time()
 7 urls=[
 8     'http://127.0.0.1:5000/bobo','http://127.0.0.1:5000/jay','http://127.0.0.1:5000/tom'
 9 ]
10 
11 async def get_page(url):
12     print('正在下载',url)
13     response=requests.get(url=url)
14     print('下载完毕',response.text)
15 
16 tasks=[]
17 
18 for url in urls:
19     c=get_page(url)
20     task=asyncio.ensure_future(c)
21     tasks.append(task)
22 
23 loop = asyncio.get_event_loop()
24 loop.run_until_complete(asyncio.wait(tasks))
25 
26 end=time.time()
27 print('总耗时:',end-start)

运行结果：

 

总耗时与串行操作差不多，为什么？

因为requests.get是基于同步的，必须使用基于异步的网络请求模块进行指定url的请求发送

而aiohttp:基于异步网络请求的模块

 1 import requests
 2 import asyncio
 3 import time
 4 import aiohttp
 5 
 6 start=time.time()
 7 urls=[
 8     'http://127.0.0.1:5000/bobo','http://127.0.0.1:5000/jay','http://127.0.0.1:5000/tom'
 9 ]
10 
11 async def get_page(url):
12     async with aiohttp.ClientSession() as session:
13         #get()、post():
14         #headers、params/data,proxy='http://ip:port'
15         async with await session.get(url) as response:
16             #text()返回字符串形式的响应数据
17             #read()返回的二进制形式的响应数据
18             #json()返回的就是json对象
19             #注意:获取响应数据操作之前一定要使用await进行手动挂起
20             page_text=await response.text()
21             print(page_text)
22     # print('正在下载',url)
23     # response=requests.get(url=url)
24     # print('下载完毕',response.text)
25 
26 tasks=[]
27 
28 for url in urls:
29     c=get_page(url)
30     task=asyncio.ensure_future(c)
31     tasks.append(task)
32 
33 loop = asyncio.get_event_loop()
34 loop.run_until_complete(asyncio.wait(tasks))
35 
36 end=time.time()
37 print('总耗时:',end-start)

运行结果：