from Day1.tuozhan_all import post, get
import json
from urllib import request, parse
# 保存cookie
from http import cookiejar
# 通过对象保存cookie
cookie_object = cookiejar.CookieJar()
# handler 对应着一个操作
handler = request.HTTPCookieProcessor(cookie_object)
# opener 遇到有cookie的response的时候,
# 调用handler内部的一个函数, 存储到cookie object
opener = request.build_opener(handler)
# def store(cookie):
# cookie_object.append(cookie)
#
# def check_response(response):
# if 'cookie' in response:
# handler.store(response['cookie'])
# url
url = 'http://www.renren.com/ajaxLogin/login?1=1&uniqueTimestamp=2018721441132'
# form
form = {
'email': '18510556963',
'icode': '',
'origURL': 'http://www.renren.com/home',
'domain': 'renren.com',
'key_id': '1',
'captcha_type': 'web_login',
'password': '95cb2a1d59b918e0d16ab5d3535fb40103e4b546e651a3e3c99b91876927c78a',
'rkey': 'a7bccfbafd7ee702247450942dff5611',
'f': 'http%3A%2F%2Fwww.renren.com%2F966927992',
}
# post
form_bytes = parse.urlencode(form).encode('utf-8')
# response = request.urlopen(url, form_bytes)
# opener = request.build_opener()
# opener.open()
response = opener.open(url, form_bytes)
html_bytes = response.read()
#html_bytes = post(url, form=form)
# 打印结果
#print(html_bytes)
# 通过json获取一个字典类型
res_dict = json.loads(html_bytes.decode('utf-8'))
home_url = res_dict['homeUrl']
# 访问页面
response = opener.open(home_url)
html_bytes = response.read()
print(html_bytes.decode('utf-8'))
----------------------------------------------------------------------------------------------------------------
Day1.tuozhan_all
from urllib import request, parse
from urllib.error import HTTPError, URLError
#a. get(url, headers=None)
def get(url, headers=None):
return urlrequests(url, headers=headers)
def post(url, form, headers=None):
return urlrequests(url, form, headers=headers)
#b. post(url, form, headers=None)
#1. 传入url
#2. user_agent
#3. headers
#4. 定义Request
#5. urlopen
#6. 返回byte数组
def urlrequests(url, form=None, headers=None):
user_agent = 'Mozilla/5.0 (Windows NT 6.1; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/67.0.3396.99 Safari/537.36'
# 如果用户需要自行传入headers, 则覆盖之前的headers
if headers == None:
headers = {
'User-Agent': user_agent
}
html_bytes = b''
try:
if form:
# POST
# 2.1 转换成str
form_str = parse.urlencode(form)
#print(form_str)
# 2.2 转换成bytes
form_bytes = form_str.encode('utf-8')
req = request.Request(url, data=form_bytes, headers=headers)
else:
# GET
req = request.Request(url, headers=headers)
response = request.urlopen(req)
html_bytes = response.read()
except HTTPError as e:
print(e)
except URLError as e:
print(e)
return html_bytes
if __name__ == '__main__':
# url = 'http://fanyi.baidu.com/sug'
# form = {
# 'kw': '呵呵'
# }
# html_bytes = post(url, form=form)
# print(html_bytes)
url = 'http://www.baidu.com'
html_byte = get(url)
print(html_byte)