应对12306反爬虫机制：JS加密参数逆向分析

// 原始请求参数
leftTicketDTO: {
  train_date: "2024-11-24",
  from_station: "BJP", 
  to_station: "SHH",
  purpose_codes: "ADULT"
}
// 实际请求URL（示例）
https://kyfw.12306.cn/otn/leftTicket/query?leftTicketDTO.train_date=2024-11-24&leftTicketDTO.from_station=BJP&leftTicketDTO.to_station=SHH&purpose_codes=ADULT&_json_att=...

// 格式化后的关键JS代码片段
function generateJsonAtt() {
    var e = Math.random().toString(36).substr(2);
    var t = new Date().getTime();
    var n = encryptMethod(e + "_" + t);
    return window.btoa(n);
}

function encryptMethod(str) {
    // 复杂的加密逻辑，可能涉及多个步骤
    var key = CryptoJS.enc.Utf8.parse('1234567812345678');
    var iv = CryptoJS.enc.Utf8.parse('1234567812345678');
    var encrypted = CryptoJS.AES.encrypt(CryptoJS.enc.Utf8.parse(str), key, {
        keySize: 128 / 8,
        iv: iv,
        mode: CryptoJS.mode.CBC,
        padding: CryptoJS.pad.Pkcs7
    });
    return encrypted.toString();
}

import execjs
import time
import random
import string

class Zhang12306JSEncrypt:
    def __init__(self):
        # 编译JS加密代码
        with open('12306_encrypt.js', 'r', encoding='utf-8') as f:
            js_code = f.read()
        self.ctx = execjs.compile(js_code)
    
    def generate_json_att(self):
        """生成_json_att参数"""
        return self.ctx.call('generateJsonAtt')

# 对应的JS文件 (12306_encrypt.js)
"""
const CryptoJS = require('crypto-js');

function generateRandomString(length) {
    return Math.random().toString(36).substr(2, length);
}

function generateJsonAtt() {
    var random_str = generateRandomString(16);
    var timestamp = new Date().getTime();
    var raw_str = random_str + "_" + timestamp;
    var encrypted = encryptMethod(raw_str);
    return Buffer.from(encrypted.toString(), 'binary').toString('base64');
}

function encryptMethod(str) {
    var key = CryptoJS.enc.Utf8.parse('1234567812345678');
    var iv = CryptoJS.enc.Utf8.parse('1234567812345678');
    var encrypted = CryptoJS.AES.encrypt(CryptoJS.enc.Utf8.parse(str), key, {
        keySize: 128 / 8,
        iv: iv,
        mode: CryptoJS.mode.CBC,
        padding: CryptoJS.pad.Pkcs7
    });
    return encrypted.ciphertext.toString(CryptoJS.enc.Base64);
}

module.exports = {
    generateJsonAtt: generateJsonAtt
};
"""

import requests
from Zhang12306JSEncrypt import Zhang12306JSEncrypt

class Advanced12306Crawler:
    def __init__(self):
        self.session = requests.Session()
        self.encryptor = Zhang12306JSEncrypt()
        
        # 代理配置
        self.proxyHost = "www.16yun.cn"
        self.proxyPort = "5445"
        self.proxyUser = "16QMSOML"
        self.proxyPass = "280651"
        
        # 设置请求头
        self.session.headers.update({
            'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36',
            'Referer': 'https://kyfw.12306.cn/otn/leftTicket/init',
        })
    
    def get_proxies(self):
        """构造代理配置字典"""
        proxy_url = f"http://{self.proxyUser}:{self.proxyPass}@{self.proxyHost}:{self.proxyPort}"
        return {
            'http': proxy_url,
            'https': proxy_url
        }
    
    def query_tickets(self, train_date, from_station, to_station):
        """查询车票信息"""
        # 生成加密参数
        json_att = self.encryptor.generate_json_att()
        
        # 构造请求参数
        params = {
            'leftTicketDTO.train_date': train_date,
            'leftTicketDTO.from_station': from_station,
            'leftTicketDTO.to_station': to_station,
            'purpose_codes': 'ADULT',
            '_json_att': json_att
        }
        
        url = 'https://kyfw.12306.cn/otn/leftTicket/query'
        
        try:
            # 使用代理发起请求
            response = self.session.get(
                url, 
                params=params, 
                proxies=self.get_proxies(),
                timeout=10  # 添加超时设置
            )
            
            if response.status_code == 200:
                data = response.json()
                if data.get('status'):
                    return self.parse_ticket_data(data['data'])
                else:
                    print("API返回状态错误:", data.get('messages', ['未知错误']))
            else:
                print(f"HTTP错误: {response.status_code}")
                
        except requests.exceptions.ProxyError as e:
            print(f"代理连接错误: {e}")
        except requests.exceptions.ConnectTimeout as e:
            print(f"连接超时: {e}")
        except requests.exceptions.ReadTimeout as e:
            print(f"读取超时: {e}")
        except Exception as e:
            print(f"请求异常: {e}")
        
        return None
    
    def parse_ticket_data(self, data):
        """解析车票数据"""
        # 简化的解析逻辑
        result = []
        for item in data.get('result', []):
            info = item.split('|')
            if len(info) > 3:
                result.append({
                    '车次': info[3],
                    '出发站': info[6],
                    '到达站': info[7],
                    '出发时间': info[8],
                    '到达时间': info[9],
                    '历时': info[10]
                })
        return result

# 使用示例
if __name__ == "__main__":
    crawler = Advanced12306Crawler()
    tickets = crawler.query_tickets('2024-11-24', 'BJP', 'SHH')
    if tickets:
        for ticket in tickets[:3]:
            print(ticket)

import hashlib
import os

class AutoUpdateEncryptor:
    def __init__(self, js_url):
        self.js_url = js_url
        self.local_file = '12306_encrypt.js'
        self.ctx = None
        self.load_js()
    
    def get_js_md5(self):
        """计算JS文件MD5以检测变化"""
        with open(self.local_file, 'rb') as f:
            return hashlib.md5(f.read()).hexdigest()
    
    def download_js(self):
        """下载最新JS文件"""
        # 实现下载逻辑
        pass
    
    def load_js(self):
        """加载JS执行环境"""
        if not os.path.exists(self.local_file):
            self.download_js()
        
        with open(self.local_file, 'r', encoding='utf-8') as f:
            js_code = f.read()
        
        self.ctx = execjs.compile(js_code)
    
    def check_and_update(self):
        """检查并更新JS文件"""
        # 定期检查JS文件是否更新
        pass