#python学习——保存网页到本地 html及pdf
直接上代码:
# -*- coding: utf-8 -*-
"""
Created on Tue Jun 14 13:01:58 2016
@author:
"""
import urllib2
import cookielib
import pdfkit
cj = cookielib.LWPCookieJar()
opener = urllib2.build_opener(urllib2.HTTPCookieProcessor(cj))
urllib2.install_opener(opener)
url = "https://www.taobao.com/"
req = urllib2.Request(url)
''' 保存html到本地'''
operate = opener.open(req)
msg = operate.read()
document = 'D://1.html'
file_ = open(document,'w')
file_.write(msg)
file_.close()
path_wk = r'C:\Python27\wkhtmltopdf\bin\wkhtmltopdf.exe'
config = pdfkit.configuration(wkhtmltopdf = path_wk)
'''保存pdf到本地'''
pdfkit.from_url(url, r'D:\are you coding\pdf\taobao.pdf', configuration=config)