import wget
import requests
import re
def getURL(url):
try:
hd={'user-agent':'Mozallia/5.0'}
r=requests.request('get',url,headers=hd)
r.raise_for_status()
r.encoding=r.apparent_encoding
html=r.text
h=re.findall(r"src=\".*.[jpg,png]\"",html)
for i in range(len(h)):
h[i]=h[i][5:-1]
return h
except:
print("爬取失败")
def download(url):
for href in url:
try:
wget.download(href,r"D:\filepython\ipython\pic\{}".format(href[-7:]))
except Exception:
continue
if __name__=="__main__":
url="http://www.tooopen.com/view/176601.html"
href=getURL(url)
download(href)