python爬虫-模拟微博登录
- 2019 年 10 月 3 日
- 筆記
微博模拟登录
这是本次爬取的网址：https://weibo.com/
一、请求分析
找到登录的位置，填写用户名密码进行登录操作
看看这次请求响应的数据是什么
这是响应得到的数据，保存下来
exectime: 8 nonce: "HW9VSX" pcid: "gz-4ede4c6269a09f5b7a6490f790b4aa944eec" pubkey: "EB2A38568661887FA180BDDB5CABD5F21C7BFD59C090CB2D245A87AC253062882729293E5506350508E7F9AA3BB77F4333231490F915F6D63C55FE2F08A49B353F444AD3993CACC02DB784ABBB8E42A9B1BBFFFB38BE18D78E87A0E41B9B8F73A928EE0CCEE1F6739884B9777E4FE9E88A1BBE495927AC4A799B3181D6442443" retcode: 0 rsakv: "1330428213" servertime: 1568257059继续完善登录操作，输入密码，点击登录按钮
经过分析呢，发现变化的参数就是sp,nonce,servetime。servetime就是当前的时间戳，我们只需找到其他两个参数的生成方法就好了。对了su这个参数是通过base64加密生成的
二、找到sp,nonce的加密方式
这次就不通过search关键字去找加密位置了
找到调用函数的位置，打上断点，再进行登录操作
经过js代码流程调试分析，最终我们找到了加密的位置
简单介绍下怎么调试js代码
找到sp,nonce的位置，通过python代码去实现它的加密方式
sp它是通过rsa加密方式，加密生成的。rsa的具体用法可以通过百度找到。或者通过sha1加密生成。至于me.rsaPubkey他怎么得到的，他就是我们还没有点击登录前，就发了一个请求，那个请求的响应数据就有它。如果你测试的次数多了的话，会发现这个值它是固定下来的。所以我们也可以直接去用，不用请求获取。
nonce：它呢也出现过在未点击登录前的那个请求响应的数据中，但是呢，我们点了几次登录，都未发现这个请求了。nonce的值每次还不一样。所以它肯定是本地js文件的某个函数生成，不用请求服务器获取。我们在这里找到了nonce的生成函数
import random import rsa import hashlib from binascii import b2a_hex def get_nonce(n): result = "" random_str = "ABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789" for i in range(n): index = random.randint(0, len(random_str) - 1) result += random_str[index] return result def get_sp_rsa(password, servertime, nonce): key = "EB2A38568661887FA180BDDB5CABD5F21C7BFD59C090CB2D245A87AC253062882729293E5506350508E7F9AA3BB77F4333231490F915F6D63C55FE2F08A49B353F444AD3993CACC02DB784ABBB8E42A9B1BBFFFB38BE18D78E87A0E41B9B8F73A928EE0CCEE1F6739884B9777E4FE9E88A1BBE495927AC4A799B3181D6442443" pubkey = rsa.PublicKey(int(key, 16), int("10001", 16)) res = rsa.encrypt(bytes("" + "t".join([servertime, nonce]) + "n" + password,encoding="utf-8"), pubkey) return b2a_hex(res) def get_sp_sha1(password, servertime, nonce): res = hashlib.sha1(bytes("" + hashlib.sha1(bytes(hashlib.sha1(bytes(password, encoding="utf-8")).hexdigest(),encoding="utf-8")).hexdigest() + servertime + nonce,encoding="utf-8")).hexdigest() return res三、响应数据
请求参数分析的差不多了，这次输入正确的用户名，密码。查看响应的数据的是什么。
打开fiddler，然后退出当前账号，重新进行登录操作。fiddler上面就会出现很多请求。找到需要的请求，看看响应内容
这样做，每个响应都会set-cookie。所以照着上面的流程实现，标识登录的cookie肯定能得到。之后的话，只要带上这个cookie去做其他操作就行了。
最后附上代码
import requests, random, time, rsa, hashlib, base64, re, json from binascii import b2a_hex class WeiBo: def __init__(self): self.session = requests.Session() self.headers = { "user-agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/76.0.3809.132 Safari/537.36" } def login(self, account, password): api = "https://login.sina.com.cn/sso/login.php?client=ssologin.js(v1.4.19)" nonce = self._get_nonce() servertime = self._get_now_time() sp = self._get_sp_rsa(password, servertime, nonce) su = self._get_su(account) data = { "entry": "weibo", "gateway": "1", "from": "", "savestate": "7", "qrcode_flag": "false", "useticket": "1", "pagerefer": "https://login.sina.com.cn/crossdomain2.php?action=logout&r=https%3A%2F%2Fpassport.weibo.com%2Fwbsso%2Flogout%3Fr%3Dhttps%253A%252F%252Fweibo.com%26returntype%3D1", "vsnf": "1", "su": su, "service": "miniblog", "servertime": servertime, "nonce": nonce, "pwencode": "rsa2", "rsakv": "1330428213", "sp": sp, "sr": "1920*1080", "encoding": "UTF - 8", "prelt": "149", "url": "https://weibo.com/ajaxlogin.php?framelogin=1&callback=parent.sinaSSOController.feedBackUrlCallBack", "returntype": "META", } headers = self.headers.copy() headers.update({ "Host": "login.sina.com.cn", "Origin": "https://weibo.com", "Referer": "https://weibo.com/" }) response = self.session.post(api, headers=headers, data=data, allow_redirects=False) search_result = self._re_search("location.replace("(.*?)"", response.text) redirct_url = search_result and search_result.group(1) if not redirct_url: raise Exception("重定向url获取失败") response = self.session.get(redirct_url, headers=headers.update({ "Referer": "https://login.sina.com.cn/sso/login.php?client=ssologin.js(v1.4.19)" }), allow_redirects=False) search_result = self._re_search('"arrURL":(.*?)}', response.text) redirct_urls = search_result and search_result.group(1) if not redirct_urls: raise Exception("重定向url获取失败") redirct_url_list = json.loads(redirct_urls) userId = "" for url in redirct_url_list: response = self.session.get(url, headers=self.headers) if url.startswith("https://passport.weibo.com/wbsso/login"): userId = self._re_search('"uniqueid":"(.*?)"', response.text).group(1) if not userId: raise Exception("userId获取失败") user_details_url = "https://weibo.com/u/{}/home?wvr=5&lf=reg".format(userId) response = self.session.get(user_details_url, headers={ "Referer": "https://weibo.com/", "user-agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/76.0.3809.132 Safari/537.36" }) if self._re_search(userId, response.text): print("登录成功") print(self.session.cookies) else: print("登录失败") def _get_nonce(self): nonce = "" random_str = "ABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789" for i in range(5): index = random.randint(0, len(random_str) - 1) nonce += random_str[index] return nonce def _get_now_time(self): return str(int(time.time())) def _get_sp_rsa(self, password, servertime, nonce): key = "EB2A38568661887FA180BDDB5CABD5F21C7BFD59C090CB2D245A87AC253062882729293E5506350508E7F9AA3BB77F4333231490F915F6D63C55FE2F08A49B353F444AD3993CACC02DB784ABBB8E42A9B1BBFFFB38BE18D78E87A0E41B9B8F73A928EE0CCEE1F6739884B9777E4FE9E88A1BBE495927AC4A799B3181D6442443" pubkey = rsa.PublicKey(int(key, 16), int("10001", 16)) res = rsa.encrypt(bytes("" + "t".join([servertime, nonce]) + "n" + password, encoding="utf-8"), pubkey) return b2a_hex(res) def _get_sp_sha1(self, password, servertime, nonce): res = hashlib.sha1(bytes("" + hashlib.sha1(bytes(hashlib.sha1(bytes(password, encoding="utf-8")).hexdigest(), encoding="utf-8")).hexdigest() + servertime + nonce, encoding="utf-8")).hexdigest() return res def _get_su(self, account): return str(base64.b64encode(bytes(account, encoding="utf-8")), encoding="utf-8") def _re_search(self, pattern, html): return re.search(pattern, html, re.S) def test(self): self.login("18716758777", "123456") if __name__ == '__main__': wb = WeiBo() wb.test() 