Files
chat_app/backend/scripts/decrypt_chat_data.py
2025-12-11 03:53:47 +08:00

390 lines
14 KiB
Python
Raw Permalink Blame History

This file contains ambiguous Unicode characters
This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.
#!/usr/bin/env python3
"""
企业微信会话存档解密脚本
使用RSA私钥解密加密的会话数据
"""
import json
import os
import sys
import base64
from pathlib import Path
from datetime import datetime
from dotenv import load_dotenv
# 加载环境变量
load_dotenv()
# 添加当前目录到路径,以便导入wework_ctypes
sys.path.insert(0, str(Path(__file__).parent))
from wework_ctypes import WeWorkFinanceSDK, WeWorkSDKError
def load_rsa_private_key(key_path: str) -> str:
"""
加载RSA私钥
Args:
key_path: 私钥文件路径
Returns:
私钥内容字符串
"""
try:
with open(key_path, 'r') as f:
key_content = f.read()
return key_content
except Exception as e:
raise Exception(f"加载RSA私钥失败: {e}")
def rsa_decrypt_base64(encrypted_data_base64: str, private_key: str) -> str:
"""
使用RSA私钥解密Base64编码的数据
Args:
encrypted_data_base64: Base64编码的加密数据
private_key: RSA私钥字符串
Returns:
解密后的字符串
"""
try:
# 使用cryptography库进行RSA解密
from cryptography.hazmat.primitives import serialization
from cryptography.hazmat.primitives.asymmetric import padding
from cryptography.hazmat.backends import default_backend
import base64
# 加载私钥
private_key_obj = serialization.load_pem_private_key(
private_key.encode('utf-8'),
password=None,
backend=default_backend()
)
# Base64解码
encrypted_data = base64.b64decode(encrypted_data_base64)
# RSA解密 (使用PKCS1v15填充)
decrypted_data = private_key_obj.decrypt(
encrypted_data,
padding.PKCS1v15()
)
return decrypted_data.decode('utf-8')
except ImportError:
raise Exception("需要安装cryptography库: pip install cryptography")
except Exception as e:
raise Exception(f"RSA解密失败: {e}")
def decrypt_chat_message(encrypt_random_key: str, encrypt_chat_msg: str,
rsa_private_key_path: str, publickey_ver: int = None) -> tuple:
"""
解密聊天消息
Args:
encrypt_random_key: 加密的随机密钥(Base64编码)
encrypt_chat_msg: 加密的聊天消息
rsa_private_key_path: RSA私钥文件路径或目录
publickey_ver: publickey_ver版本号,用于动态匹配私钥
Returns:
(解密后的消息字典, 实际使用的私钥路径)
"""
print(f"正在解密消息...")
print(f"加密密钥长度: {len(encrypt_random_key)}")
print(f"加密消息长度: {len(encrypt_chat_msg)}")
# 动态确定私钥路径
actual_key_path = rsa_private_key_path
# 检查是否使用默认私钥路径(用户未指定自定义路径)
# 通过检查路径是否指向默认文件来判断
default_key_path = Path("cert/1_rsa.key")
specified_key_path = Path(rsa_private_key_path)
# 如果使用默认私钥路径且提供了publickey_ver,则动态匹配
if publickey_ver is not None and specified_key_path == default_key_path:
# 构建动态私钥路径:cert/{publickey_ver}_rsa.key
dynamic_key_path = f"cert/{publickey_ver}_rsa.key"
dynamic_path = Path(dynamic_key_path)
if dynamic_path.exists():
actual_key_path = dynamic_key_path
print(f"检测到publickey_ver={publickey_ver},使用动态私钥: {actual_key_path}")
else:
print(f"警告: 动态私钥文件不存在: {dynamic_key_path},使用默认私钥: {actual_key_path}")
# 1. 加载RSA私钥
print(f"加载RSA私钥: {actual_key_path}")
private_key = load_rsa_private_key(actual_key_path)
print(f"✓ RSA私钥加载成功")
# 2. 使用RSA私钥解密encrypt_random_key得到encrypt_key
print("正在使用RSA解密随机密钥...")
try:
encrypt_key = rsa_decrypt_base64(encrypt_random_key, private_key)
print(f"✓ RSA解密成功,encrypt_key长度: {len(encrypt_key)}")
except Exception as e:
print(f"✗ RSA解密失败: {e}")
# 尝试直接使用encrypt_random_key作为encrypt_key(某些情况下可能已经解密)
print("尝试使用原始encrypt_random_key作为encrypt_key...")
encrypt_key = encrypt_random_key
# 3. 使用SDK解密聊天消息
print("正在使用SDK解密聊天消息...")
try:
# 初始化SDK(只需要解密功能,不需要corpid和secret)
with WeWorkFinanceSDK("dummy", "dummy") as sdk:
decrypted_msg = sdk.decrypt_data(encrypt_key, encrypt_chat_msg)
print("✓ SDK解密成功")
# 尝试解析JSON
try:
result = json.loads(decrypted_msg)
return result, actual_key_path
except json.JSONDecodeError:
# 如果不是JSON,返回原始文本
return {"raw_message": decrypted_msg}, actual_key_path
except WeWorkSDKError as e:
raise Exception(f"SDK解密失败: {e}")
except Exception as e:
raise Exception(f"解密过程出错: {e}")
def decrypt_file(input_file: str, rsa_private_key_path: str, output_file: str = None):
"""
解密文件中的聊天数据
Args:
input_file: 输入文件路径(JSON格式)
rsa_private_key_path: RSA私钥文件路径
output_file: 输出文件路径(可选)
"""
print("=" * 60)
print("企业微信会话存档解密脚本")
print("=" * 60)
# 检查输入文件
input_path = Path(input_file)
if not input_path.exists():
print(f"错误: 输入文件不存在: {input_file}")
return 1
# 检查RSA私钥文件(只检查用户指定的路径,动态路径在解密时检查)
rsa_path = Path(rsa_private_key_path)
if not rsa_path.exists():
print(f"错误: 指定的RSA私钥文件不存在: {rsa_private_key_path}")
return 1
print(f"输入文件: {input_path.absolute()}")
print(f"指定私钥: {rsa_path.absolute()}")
print("-" * 60)
# 用于记录实际使用的私钥路径
actual_key_paths_used = set()
try:
# 读取输入文件
with open(input_path, 'r', encoding='utf-8') as f:
data = json.load(f)
print(f"文件加载成功")
# 检查数据格式
if "chatdata" not in data:
print("错误: 文件中没有chatdata字段")
return 1
chatdata = data.get("chatdata", [])
chat_count = len(chatdata)
print(f"共发现 {chat_count} 条聊天记录")
if chat_count == 0:
print("没有可解密的聊天记录")
return 0
# 解密所有消息
decrypted_messages = []
success_count = 0
fail_count = 0
for i, msg in enumerate(chatdata):
# print('msg', type(msg))
seq = msg.get("seq", i + 1)
msgid = msg.get("msgid", f"msg_{i}")
encrypt_random_key = msg.get("encrypt_random_key", "")
encrypt_chat_msg = msg.get("encrypt_chat_msg", "")
print(f"\n[{i+1}/{chat_count}] 解密消息 seq={seq}, msgid={msgid[:20]}...")
if not encrypt_random_key or not encrypt_chat_msg:
print(f" 跳过: 缺少加密数据")
fail_count += 1
decrypted_messages.append({
"seq": seq,
"msgid": msgid,
"error": "缺少加密数据",
"original": msg
})
continue
try:
# 获取publickey_ver
publickey_ver = msg.get("publickey_ver")
# 解密消息
decrypted_msg, actual_key_path = decrypt_chat_message(
encrypt_random_key,
encrypt_chat_msg,
rsa_private_key_path,
publickey_ver
)
# 记录实际使用的私钥路径
actual_key_paths_used.add(actual_key_path)
# 添加元数据
if isinstance(decrypted_msg, dict):
decrypted_msg["_metadata"] = {
"seq": seq,
"msgid": msgid,
"decrypt_time": datetime.now().isoformat(),
"original_has_encryption": True
}
decrypted_messages.append({
"seq": seq,
"msgid": msgid,
"decrypted": decrypted_msg,
"original_summary": {
"publickey_ver": msg.get("publickey_ver"),
"encrypt_key_len": len(encrypt_random_key),
"encrypt_msg_len": len(encrypt_chat_msg)
}
})
success_count += 1
print(f" ✓ 解密成功")
# 显示解密后的消息摘要
if isinstance(decrypted_msg, dict):
msg_type = decrypted_msg.get("msgtype", "unknown")
from_user = decrypted_msg.get("from", "unknown")
to_user = decrypted_msg.get("tolist", ['unknown'])[0]
if isinstance(decrypted_msg.get("tolist"), list) and len(decrypted_msg.get("tolist")) > 1:
to_user += f" 等 {len(decrypted_msg.get('tolist'))} 人"
print(f" 类型: {msg_type}, 来自: {from_user}, 发送给: {to_user}")
# 成功解密后继续处理下一条消息
continue
except Exception as e:
print(f" ✗ 解密失败: {e}")
fail_count += 1
decrypted_messages.append({
"seq": seq,
"msgid": msgid,
"error": str(e),
"original": msg
})
print("\n" + "-" * 60)
print(f"解密完成: 成功 {success_count} 条, 失败 {fail_count} 条")
# 显示实际使用的私钥信息
if actual_key_paths_used:
print(f"实际使用的私钥文件:")
for key_path in sorted(actual_key_paths_used):
print(f" - {key_path}")
# 生成输出文件名
if output_file is None:
timestamp = datetime.now().strftime("%y%m%d_%H%M%S")
output_file = f"data/decrypted_{input_path.stem}_{timestamp}.json"
output_path = Path(output_file)
# 准备输出数据
output_data = {
"metadata": {
"input_file": str(input_path.absolute()),
"decrypt_time": datetime.now().isoformat(),
"total_messages": chat_count,
"successful_decrypts": success_count,
"failed_decrypts": fail_count,
"specified_key": str(rsa_path.absolute()),
"actual_keys_used": sorted([str(Path(p).absolute()) for p in actual_key_paths_used])
},
"decrypted_messages": decrypted_messages,
"original_metadata": data.get("_metadata", {})
}
# 保存解密结果
with open(output_path, 'w', encoding='utf-8') as f:
json.dump(output_data, f, indent=2, ensure_ascii=False)
print(f"解密结果已保存到: {output_path.absolute()}")
# 显示一些解密后的消息示例
if success_count > 0:
print("\n解密消息示例:")
for i, msg in enumerate(decrypted_messages[:3]): # 显示前3条
if "decrypted" in msg:
decrypted = msg["decrypted"]
if isinstance(decrypted, dict):
seq = msg.get("seq", "N/A")
msg_type = decrypted.get("msgtype", "unknown")
content = decrypted.get("content", "无内容")[:100]
print(f" [{i+1}] seq={seq}, 类型={msg_type}")
print(f" 内容: {content}...")
print("\n" + "=" * 60)
print("解密完成")
return 0
except json.JSONDecodeError as e:
print(f"JSON解析错误: {e}")
return 1
except Exception as e:
print(f"解密过程出错: {e}")
import traceback
traceback.print_exc()
return 1
def main():
"""主函数"""
import argparse
parser = argparse.ArgumentParser(
description="企业微信会话存档解密脚本",
formatter_class=argparse.RawDescriptionHelpFormatter,
epilog="""
使用示例:
# 解密data目录下的文件
%(prog)s data/chat_0_100_251203.json
# 指定RSA私钥路径
%(prog)s data/chat_0_100_251203.json --key /path/to/private_key.pem
# 指定输出文件
%(prog)s data/chat_0_100_251203.json --output decrypted_result.json
"""
)
parser.add_argument('input_file', help='输入文件路径(JSON格式)')
parser.add_argument('--key', '-k', default='cert/1_rsa.key',
help='RSA私钥文件路径,默认: cert/1_rsa.key')
parser.add_argument('--output', '-o', help='输出文件路径,默认自动生成')
args = parser.parse_args()
return decrypt_file(args.input_file, args.key, args.output)
if __name__ == "__main__":
sys.exit(main())