【python】爬取中国大学排名 

以下代码为爬取2015年中国大学排名及年费
 1 import requests
 2 from bs4 import BeautifulSoup
 3 allUniv = []
 4 def getHTMLText(url):
 5     try:
 6         r = requests.get(url, timeout=30)
 7         r.raise_for_status()
 8         r.encoding = 'utf-8'
 9         return r.text
10     except:
11         return ""
12 def fillUnivList(soup):
13     data = soup.find_all('tr')
14     for tr in data:
15         ltd = tr.find_all('td')
16         if len(ltd)==0:
17             continue
18         singleUniv = []
19         for td in ltd:
20             singleUniv.append(td.string)
21         allUniv.append(singleUniv)
22 def printUnivList(num):
23     print("{1:^2}{2:{0}^10}{3:{0}^6}{4:{0}^4}{5:{0}^10}".format(chr(12288),"排名","学校名称","省市","总分","年费"))
24     for i in range(num):
25         u=allUniv[i]
26         print("{1:^4}{2:{0}^10}{3:{0}^5}{4:{0}^8.1f}{5:{0}^11}".format(chr(12288),u[0],u[1],u[2],eval(u[3]),u[4]))
27 def main():
28     url = 'http://www.zuihaodaxue.com/zuihaodaxuepaiming2015_3.html'
29     html = getHTMLText(url)
30     soup = BeautifulSoup(html, "html.parser")
31     fillUnivList(soup)
32     printUnivList(10)
33 main()

 

posted @ 2020-05-15 22:24  litchi31  阅读(313)  评论(0)    收藏  举报