微博用户评论爬取
程序员文章站
2022-05-02 20:42:54
...
# *第一部分首先要爬取MiuMiu的每一条微博的mid，mid就是每一条微博的唯一标识符，便于后期直接爬取；
# 此次爬取下来的数据是：mid和评论数两个维度，后去需要将没有评论的mid删除，节省爬虫时间。*
#-------------------------------------------------------------------------------------#
#使用selenium模块进行模拟浏览器爬取，在python下直接pip install selenium安装即可；
#另外需要安装chrome或者Firefox浏览器，推荐使用安装chrome浏览器，并且要根据chrome浏览器的相应版本安装chromedriver，
#chromedriver的版本必须要与安装的chrome版本相对应，要不然selenium调浏览器会出错。
from selenium import webdriver
import time
driver = webdriver.Chrome()
driver.get('https://weibo.com')
time.sleep(5)
driver.find_element_by_id('loginName').send_keys('******')
driver.find_element_by_id('loginPassword').send_keys('******')
driver.find_element_by_id('loginAction').click()
time.sleep(5)
driver.get('https://weibo.com/miumiuofficial?profile_ftype=1&is_all=1#_0')#此网址为你需要爬取的微博的主页，手动复制添加即可
for j in range(29):
print(j)
for i in range(3):
driver.execute_script('window.scrollTo(0,1000000)')
time.sleep(10)
for i in range(2,47):
test = driver.find_element_by_xpath('//*[@id="Pl_Official_MyProfileFeed__23"]/div/div['+str(i)+']/div[1]/div[3]/div[2]/a[1]')
comment = driver.find_element_by_xpath('//*[@id="Pl_Official_MyProfileFeed__23"]/div/div['+str(i)+']/div[2]/div/ul/li[3]/a/span/span/span/em[2]').text
url_mid = test.get_attribute('name')
print(i)
with open('D:\\newmid_MiuMiu.txt',mode='a+') as f:#数据保存地址
f.write(url_mid)
f.write('\t')
f.write(str(comment))
f.write('\n')
driver.find_element_by_link_text('下一页').click()
time.sleep(10)
#第二部分进行评论数据爬取，爬取的格式为：评论 时间两个维度进行保存。#
#-------------------------------------------------------------#
# -*- coding: utf-8 -*-
import time
import base64
import rsa
import binascii
import requests
import re
from PIL import Image
import random
from urllib.parse import quote_plus
import http.cookiejar as cookielib
import json
import pandas as pd
import pandas
agent = 'mozilla/5.0 (windowS NT 10.0; win64; x64) appLewEbkit/537.36 (KHTML, likE gecko) chrome/71.0.3578.98 safari/537.36'
headers = {'User-Agent': agent}
class WeiboLogin(object):
"""
通过登录 weibo.com 然后跳转到 m.weibo.cn
"""
#初始化数据
def __init__(self, user, password, cookie_path):
super(WeiboLogin, self).__init__()
self.user = user
self.password = password
self.session = requests.Session()
self.cookie_path = cookie_path
# LWPCookieJar是python中管理cookie的工具，可以将cookie保存到文件，或者在文件中读取cookie数据到程序
self.session.cookies = cookielib.LWPCookieJar(filename=self.cookie_path)
self.index_url = "http://weibo.com/login.php"
self.session.get(self.index_url, headers=headers, timeout=2)
self.postdata = dict()
def get_su(self):
"""
对 email 地址和手机号码 先 javascript 中 encodeURIComponent
对应 Python 3 中的是 urllib.parse.quote_plus
然后在 base64 加密后decode
"""
username_quote = quote_plus(self.user)
username_base64 = base64.b64encode(username_quote.encode("utf-8"))
return username_base64.decode("utf-8")
# 预登陆获得 servertime, nonce, pubkey, rsakv
def get_server_data(self, su):
"""与原来的相比，微博的登录从 v1.4.18 升级到了 v1.4.19
这里使用了 URL 拼接的方式，也可以用 Params 参数传递的方式
"""
pre_url = "http://login.sina.com.cn/sso/prelogin.php?entry=weibo&callback=sinaSSOController.preloginCallBack&su="
pre_url = pre_url + su + "&rsakt=mod&checkpin=1&client=ssologin.js(v1.4.19)&_="
pre_url = pre_url + str(int(time.time() * 1000))
pre_data_res = self.session.get(pre_url, headers=headers)
sever_data = eval(pre_data_res.content.decode("utf-8").replace("sinaSSOController.preloginCallBack", ''))
return sever_data
def get_password(self, servertime, nonce, pubkey):
"""对密码进行 RSA 的加密"""
rsaPublickey = int(pubkey, 16)
key = rsa.PublicKey(rsaPublickey, 65537) # 创建公钥
message = str(servertime) + '\t' + str(nonce) + '\n' + str(self.password) # 拼接明文js加密文件中得到
message = message.encode("utf-8")
passwd = rsa.encrypt(message, key) # 加密
passwd = binascii.b2a_hex(passwd) # 将加密信息转换为16进制。
return passwd
def get_cha(self, pcid):
"""获取验证码，并且用PIL打开，
1. 如果本机安装了图片查看软件，也可以用 os.subprocess 的打开验证码
2. 可以改写此函数接入打码平台。
"""
cha_url = "https://login.sina.com.cn/cgi/pin.php?r="
cha_url = cha_url + str(int(random.random() * 100000000)) + "&s=0&p="
cha_url = cha_url + pcid
cha_page = self.session.get(cha_url, headers=headers)
with open("cha.jpg", 'wb') as f:
f.write(cha_page.content)
f.close()
try:
im = Image.open("cha.jpg")
im.show()
im.close()
except Exception as e:
print(u"请到当前目录下，找到验证码后输入")
def pre_login(self):
# su 是加密后的用户名
su = self.get_su()
sever_data = self.get_server_data(su)
servertime = sever_data["servertime"]
nonce = sever_data['nonce']
rsakv = sever_data["rsakv"]
pubkey = sever_data["pubkey"]
showpin = sever_data["showpin"] # 这个参数的意义待探索
password_secret = self.get_password(servertime, nonce, pubkey)
self.postdata = {
'entry': 'weibo',
'gateway': '1',
'from': '',
'savestate': '7',
'useticket': '1',
'pagerefer': "https://passport.weibo.com",
'vsnf': '1',
'su': su,
'service': 'miniblog',
'servertime': servertime,
'nonce': nonce,
'pwencode': 'rsa2',
'rsakv': rsakv,
'sp': password_secret,
'sr': '1366*768',
'encoding': 'UTF-8',
'prelt': '115',
"cdult": "38",
'url': 'http://weibo.com/ajaxlogin.php?framelogin=1&callback=parent.sinaSSOController.feedBackUrlCallBack',
'returntype': 'TEXT' # 这里是 TEXT 和 META 选择，具体含义待探索
}
return sever_data
def login(self):
# 先不输入验证码登录测试
try:
sever_data = self.pre_login()
login_url = 'https://login.sina.com.cn/sso/login.php?client=ssologin.js(v1.4.19)&_'
login_url = login_url + str(time.time() * 1000)
login_page = self.session.post(login_url, data=self.postdata, headers=headers)
ticket_js = login_page.json()
ticket = ticket_js["ticket"]
except Exception as e:
sever_data = self.pre_login()
login_url = 'https://login.sina.com.cn/sso/login.php?client=ssologin.js(v1.4.19)&_'
login_url = login_url + str(time.time() * 1000)
pcid = sever_data["pcid"]
self.get_cha(pcid)
self.postdata['door'] = input(u"请输入验证码")
login_page = self.session.post(login_url, data=self.postdata, headers=headers)
ticket_js = login_page.json()
ticket = ticket_js["ticket"]
# 以下内容是 处理登录跳转链接
save_pa = r'==-(\d+)-'
ssosavestate = int(re.findall(save_pa, ticket)[0]) + 3600 * 7
jump_ticket_params = {
"callback": "sinaSSOController.callbackLoginStatus",
"ticket": ticket,
"ssosavestate": str(ssosavestate),
"client": "ssologin.js(v1.4.19)",
"_": str(time.time() * 1000),
}
jump_url = "https://passport.weibo.com/wbsso/login"
jump_headers = {
"Host": "passport.weibo.com",
"Referer": "https://weibo.com/",
"User-Agent": headers["User-Agent"]
}
jump_login = self.session.get(jump_url, params=jump_ticket_params, headers=jump_headers)
uuid = jump_login.text
uuid_pa = r'"uniqueid":"(.*?)"'
uuid_res = re.findall(uuid_pa, uuid, re.S)[0]
web_weibo_url = "http://weibo.com/%s/profile?topnav=1&wvr=6&is_all=1" % uuid_res
weibo_page = self.session.get(web_weibo_url, headers=headers)
Mheaders = {
"Host": "login.sina.com.cn",
"User-Agent": agent
}
# m.weibo.cn 登录的 url 拼接
_rand = str(time.time())
mParams = {
"url": "https://m.weibo.cn/",
"_rand": _rand,
"gateway": "1",
"service": "sinawap",
"entry": "sinawap",
"useticket": "1",
"returntype": "META",
"sudaref": "",
"_client_version": "0.6.26",
}
murl = "https://login.sina.com.cn/sso/login.php"
mhtml = self.session.get(murl, params=mParams, headers=Mheaders)
mhtml.encoding = mhtml.apparent_encoding
mpa = r'replace\((.*?)\);'
mres = re.findall(mpa, mhtml.text)
# 关键的跳转步骤，这里不出问题，基本就成功了。
Mheaders["Host"] = "passport.weibo.cn"
self.session.get(eval(mres[0]), headers=Mheaders)
mlogin = self.session.get(eval(mres[0]), headers=Mheaders)
# print(mlogin.status_code)
# 进过几次 页面跳转后，m.weibo.cn 登录成功，下次测试是否登录成功
Mheaders["Host"] = "m.weibo.cn"
Set_url = "https://m.weibo.cn"
pro = self.session.get(Set_url, headers=Mheaders)
pa_login = r'isLogin":true,'
login_res = re.findall(pa_login, pro.text)
# 可以通过 session.cookies 对 cookies 进行下一步相关操作
self.session.cookies.save()
def weibo_comment(id_mid):
max_id = ""
headers = {"user-agent": "mozilla/5.0 (windowS NT 10.0; win64; x64) appLewEbkit/537.36 (KHTML, likE gecko) chrome/71.0.3578.98 safari/537.36"}
#加载cookie
cookies = cookielib.LWPCookieJar("Cookie.txt")
cookies.load(ignore_discard=True, ignore_expires=True)
# 将cookie转换成字典
cookie_dict = requests.utils.dict_from_cookiejar(cookies)
max_id_list = []
while True:
if max_id == "":
url = "https://m.weibo.cn/comments/hotflow?id="+str(id_mid)+"&mid="+str(id_mid)+"&max_id_type=0"
else:
url = "https://m.weibo.cn/comments/hotflow?id="+str(id_mid)+"&mid="+str(id_mid)+"&max_id="+str(max_id)+"&max_id_type=0"
response = requests.get(url, headers=headers, cookies=cookie_dict)
comment = response.json()
if comment['ok'] == 0:
break
max_id = comment["data"]["max_id"]
if max_id in max_id_list:
break
max_id_list.append(max_id)
line = []
for comment_data in comment["data"]["data"]:
data = comment_data["text"]
p = re.compile(r'(<span.*>.*</span>)*(<a.*>.*</ a>)?')
data = p.sub(r'', data)
data1 = [comment_data["text"]]
data2 = pandas.DataFrame(data1,columns=["text"])
data2["date"] = comment_data["created_at"]
data2.to_csv('D:\\MiuMiu.csv', header=False,index=False,mode='a+',encoding="utf_8_sig") #写入csv文件,'a+'是追加模式
if len(data) != 0:
line.append(data)
time.sleep(5)
if __name__ == '__main__':
username = "******" # 用户名
password = "******" # 密码
cookie_path = "Cookie.txt" # 保存cookie 的文件名称
weibo = WeiboLogin(username, password, cookie_path)
weibo.login()
data_MiuMiu = pd.read_table("D:\\MiuMiu.txt",header=None,sep=',')
data_MiuMiu_1 = data_MiuMiu[0]
num_loop = len(data_MiuMiu_1)
for i in range(88,num_loop):
print(i)
weibo_comment(str(data_MiuMiu_1[i]))
下一篇： 爬取新浪微博评论