asyncio是从pytohn3.4开始添加到标准库中的一个强大的异步并发库，可以很好地解决python中高并发的问题，入门学习可以参考官方文档
并发访问能极大的提高爬虫的性能，但是requests访问网页是阻塞的，无法并发，所以我们需要一个更牛逼的库 aiohttp ，它的用法与requests相似，可以看成是异步版的requests，下面通过实战爬取猫眼电影专业版来熟悉它们的使用：
1. 分析
分析网页源代码发现猫眼专业版是一个动态网页，其中的数据都是后台传送的，打开F12调试工具，再刷新网页选择XHR后发现第一条就是后台发来的电影数据，由此得到接口 https://box.maoyan.com/promovie/api/box/second.json?beginDate=日期
在这里插入图片描述
2. 异步爬取
创建20个任务来并发爬取20天的电影信息并写入csv文件，同时计算一下耗费的时间
import asyncio
from aiohttp import ClientSession
import time
import csv headers = {'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) ' 'AppleWebKit/537.36 (KHTML, like Gecko) ' 'Chrome/67.0.3396.99 Safari/537.36'} # 协程函数，完成一个无阻塞的任务 async def get_one_page(url): try: async with ClientSession() as session: async with session.get(url, headers=headers) as r: # 返回解析为字典的电影数据 return await r.json() except Exception as e: print('请求异常： ' + str(e)) return {} # 解析函数，提取每一条内容并写入csv文件 def parse_one_page(movie_dict, writer): try: movie_list = movie_dict['data']['list'] for movie in movie_list: movie_name = movie['movieName'] release = movie['releaseInfo'] sum_box = movie['sumBoxInfo'] box_info = movie['boxInfo'] box_rate = movie['boxRate'] show_info = movie['showInfo'] show_rate = movie['showRate'] avg_show_view = movie['avgShowView'] avg_seat_view = movie['avgSeatView'] writer.writerow([movie_name, release, sum_box, box_info, box_rate, show_info, show_rate, avg_show_view, avg_seat_view]) return('写入成功') except Exception as e: return('解析异常： ' + str(e)) # 并发爬取 async def main(): # 待访问的20个URL链接 urls = ['https://box.maoyan.com/promovie/api/box/second.json?beginDate=201904{}{}'.format(i, j) for i in range(1, 3) for j in range(10)] # 任务列表 tasks = [get_one_page(url) for url in urls] # 并发执行并保存每一个任务的返回结果 results = await asyncio.gather(*tasks) # 处理每一个结果 with open('out/pro_info.csv', 'w') as f: writer = csv.writer(f) for result in results: print(parse_one_page(result, writer)) if __name__ == "__main__": start = time.time() asyncio.run(main()) # python3.7之前的写法 # loop = asyncio.get_event_loop() # loop.run_until_complete(main()) # loop.close() print(time.time()-start) 
在这里插入图片描述
用时0.15秒！
3. 对比同步爬取
import requests
import csv
import time
headers = {'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) ' 'AppleWebKit/537.36 (KHTML, like Gecko) ' 'Chrome/67.0.3396.99 Safari/537.36'} def get_one_page(url): try: r = requests.get(url, headers=headers) r.raise_for_status() r.encoding = r.apparent_encoding return r.json() except Exception as e: print('请求异常： ' + e) return {} def parse_one_page(movie_dict, writer): try: movie_list = movie_dict['data']['list'] for movie in movie_list: movie_name = movie['movieName'] release = movie['releaseInfo'] sum_box = movie['sumBoxInfo'] box_info = movie['boxInfo'] box_rate = movie['boxRate'] show_info = movie['showInfo'] show_rate = movie['showRate'] avg_show_view = movie['avgShowView'] avg_seat_view = movie['avgSeatView'] writer.writerow([movie_name, release, sum_box, box_info, box_rate, show_info, show_rate, avg_show_view, avg_seat_view]) print('写入成功') except Exception as e: print('解析异常： ' + e) def main(): # 待访问的20个URL链接 urls = ['https://box.maoyan.com/promovie/api/box/second.json?beginDate=201903{}{}'.format(i, j) for i in range(1, 3) for j in range(10)] with open('out/pro_info.csv', 'w') as f: writer = csv.writer(f) for url in urls: # 逐一处理 movie_dict = get_one_page(url) parse_one_page(movie_dict, writer) if __name__ == '__main__': a = time.time() main() print(time.time() - a) 
在这里插入图片描述
可以看到使用asyncio+aiohttp的异步爬取方式要比简单的requests同步爬取快上不少，尤其是爬取大量网页的时候，这种差距会非常明显。