代码改变世界

selenium加载本地配置文件 ; 获取页面动态html ; 静态html

2016-05-03 11:49  LI桥IL  阅读(1548)  评论(0编辑  收藏  举报
 1 # -*- coding: utf-8 -*-
 2 from selenium import webdriver
 3 import sys
 4 import urllib2
 5 import time
 6 import codecs
 7 reload(sys)
 8 sys.setdefaultencoding('utf-8')
 9 
10 def gethtml(url):
11     page = urllib2.urlopen(url)
12     html_all = page.read()
13     return html_all
14 
15 def get_jingtai_dongtai_html(url):
16     fp=webdriver.FirefoxProfile(r'C:\Users\j\AppData\Roaming\Mozilla\Firefox\Profiles\pn4ffyp2.default')
17     sel = webdriver.Firefox(fp)
18     sel.get(url)
19     time.sleep(3)
20     sel.set_window_size(480, 600)
21     html1=sel.page_source
22     with codecs.open('result-dongtai4.html','w',encoding='utf-8')as putin1:
23         putin1.write(html1)
24     html2=gethtml(url)
25     with codecs.open('result-jingtai4.html','w',encoding='utf-8')as putin2:
26         putin2.write(html2)
27     sel.close()
28     sel.quit()
29 url="http://patent.ip1840.com/faming/186246.html"
30 get_jingtai_dongtai_html(url)