#!/usr/bin/python
import re
import urllib
import urllib.request #python3中urlopen、urlritrieve都在request庫裏面了,所以要導入此庫
def htmlGet(url):
page = urllib.request.urlopen(url)
html = page.read()
return html
def imgGet(html):
res = r'src="(https.*?.jpg)"'
imgre = re.compile(res)
imglist = re.findall(imgre,html.decode("utf-8")) #html不加後面的會報錯typeerror,因爲編碼格式的變化,這裏需要指定一下
x = 0
for i in imglist:
urllib.request.urlretrieve(i,"%s.jpg" % x)
x+=1
html = htmlGet("http://***")
imgGet(html)