前言
Python是个获取数据的小能手,所以这次希望能用它在*乎爬取一些的问题的回答数,练练手。
1.导入模块
- import re
- from bs4 import BeautifulSoup
- import requests
- import time
- import json
- import pandas as pd
- import numpy as np
2.状态码
- r = requests.get('https://github.com/explore')
- r.status_code
3. 爬取*乎
- #浏览器header和cookies
- headers = {'User-Agent':'Mozilla/5.0 (Macintosh; Intel Mac OS X 10_14_6) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/80.0.3987.87 Safari/537.36'}
- cookies = {'cookie':'_zap=3d979dbb-f25b-4014-8770-89045dec48f6; d_c0="APDvML4koQ-PTqFU56egNZNd2wd-eileT3E=|1561292196"; tst=r; _ga=GA1.2.910277933.1582789012; q_c1=9a429b07b08a4ae1afe0a99386626304|1584073146000|1561373910000; _xsrf=bf1c5edf-75bd-4512-8319-02c650b7ad2c; _gid=GA1.2.1983259099.1586575835; l_n_c=1; l_cap_id="NDIxM2M4OWY4N2YwNDRjM2E3ODAxMDdmYmY2NGFiMTQ=|1586663749|ceda775ba80ff485b63943e0baf9968684237435"; r_cap_id="OWY3OGQ1MDJhMjFjNDBiYzk0MDMxMmVlZDIwNzU0NzU=|1586663749|0948d23c731a8fa985614d3ed58edb6405303e99"; cap_id="M2I5NmJkMzRjMjc3NGZjNDhiNzBmNDMyNDQ3NDlmNmE=|1586663749|dacf440ab7ad64214a939974e539f9b86ddb9eac"; n_c=1; Hm_lvt_98beee57fd2ef70ccdd5ca52b9740c49=1586585625,1586587735,1586667228,1586667292; Hm_lpvt_98beee57fd2ef70ccdd5ca52b9740c49=1586667292; SESSIONID=GWBltmMTwz5oFeBTjRm4Akv8pFF6p8Y6qWkgUP4tjp6; JOID=UVkSBEJI6EKgHAipMkwAEWAkvEomDbkAwmJn4mY1kHHPVGfpYMxO3voUDK88UO62JqgwW5Up4hC2kX_KGO9xoKI=; osd=UlEXAU5L4EelEAuhN0kMEmghuUYlBbwFzmFv52M5k3nKUWvqaMlL0vkcCaowU-azI6QzU5As7hO-lHrGG-d0pa4=; capsion_ticket="2|1:0|10:1586667673|14:capsion_ticket|44:YTJkYmIyN2Q4YWI4NDI0Mzk0NjQ1YmIwYmUxZGYyNzY=|b49eb8176314b73e0ade9f19dae4b463fb970c8cbd1e6a07a6a0e535c0ab8ac3"; z_c0="2|1:0|10:1586667694|4:z_c0|92:Mi4xOGc1X0dnQUFBQUFBOE84d3ZpU2hEeVlBQUFCZ0FsVk5ydTVfWHdDazlHMVM1eFU5QjlqamJxWVhvZ2xuWlhTaVJ3|bcd3601ae34951fe72fd3ffa359bcb4acd60462715edcd1e6c4e99776f9543b3"; unlock_ticket="AMCRYboJGhEmAAAAYAJVTbankl4i-Y7Pzkta0e4momKdPG3NRc6GUQ=="; KLBRSID=fb3eda1aa35a9ed9f88f346a7a3ebe83|1586667697|1586660346'}
- start_url = 'https://www.zhihu.com/api/v3/feed/topstory/recommend?session_token=c03069ed8f250472b687fd1ee704dd5b&desktop=true&page_number=5&limit=6&action=pull&ad_interval=-1&before_id=23'
4. beautifulsoup解析
- s = requests.Session()
- start_url = 'https://www.zhihu.com/'
- html = s.get(url = start_url, headers = headers,cookies = cookies,timeout = 5)
- soup = BeautifulSoup(html.content)
- question = [] ## 名称
- question_address = [] ## url
- temp1 = soup.find_all('div',class_='Card TopstoryItem TopstoryItem-isRecommend')
- for item in temp1:
- temp2 = item.find_all('div',itemprop="zhihu:question")
- # print(temp2)
- if temp2 != []: #### 存在专栏等情况,暂时跳过
- question_address.append(temp2[0].find('meta',itemprop='url').get('content'))
- question.append(temp2[0].find('meta',itemprop='name').get('content'))
5. 存储信息
- question_focus_number = [] #关注量
- question_answer_number = [] # 回答量
- for url in question_address:
- test = s.get(url = url,headers = headers,cookies = cookies,timeout = 5)
- soup = BeautifulSoup(test.content)
- info = soup.find_all('div',class_='QuestionPage')[0]
- # print(info)
- focus_number = info.find('meta',itemprop="answerCount").get('content')
- answer_number = info.find('meta',itemprop="zhihu:followerCount").get('content')
- question_focus_number.append(focus_number)
- question_answer_number.append(answer_number)
6. 整理信息并输出
- question_info = pd.DataFrame(list(zip(question,question_focus_number,question_answer_number)),columns = ['问题名称','关注人数','回答人数']
- for item in ['关注人数','回答人数']:
- question_info[item] = np.array(question_info[item],dtype = 'int')
- question_info.sort_values(by='关注人数',ascending = False)
输出:
7. 总计:
简单的爬取并不难,但涉及到账户密码等,就需要注意了。爬取数据尽量不要给人家服务器造成负担(比如:把睡眠时间加长);不要把爬取的数据用于商业行为;不管技术有多牛,不要轻易触碰用户隐私数据。合理、合法、有节制的利用爬虫技术,要不可能给自己带来不必要的麻烦。