该楼层疑似违规已被系统折叠 隐藏此楼查看此楼
import requests
from requests import Request,Session
headers={
'User-Agent':"Mozilla/5.0 (Windows NT 6.1; WOW64)"
"AppleWebKit/537.36 (KHTML, like Gecko) Chrome/39.0.2171.71 Safari/537.36"
}
def loadpage(fullurl,filename):
print("正在下载:",filename)
s = Session()
request = Request('GET', fullurl)
prepped = s.prepare_request(request)
response = s.send(prepped)
return response
def writepage(html,filename):
print("正在保存:",filename)
with open(filename,"wb") as f:
f.write(html)
def get_url(url,beginpage,endpage):
for page in range(beginpage,endpage+1):
fullurl=url+"&pn="+str((page-1)*50)
filename="C:/第"+str(page)+"页.html"
html=loadpage(fullurl,filename)
writepage(html,filename)
if __name__=="_main_":
kw=input("请输入贴吧名:")
beginpage=int(input("请输入起始页:"))
endpage=int(input("请输入结束页:"))
url="http://tieba.baidu.com/f?&ie=utf-8&kw="
url=url+kw
get_url(url,beginpage,endpage)