Python数据脱敏案例如何隐藏隐私数据

wen python案例 2026-06-29 235

本文目录导读：

常用脱敏工具库
基础脱敏函数
实际应用案例
使用Faker生成假数据替换
批量数据脱敏处理
高级脱敏模式（差分隐私）
正则表达式智能脱敏
完整的数据脱敏流程
使用建议

我来为你展示几个Python数据脱敏的实用案例：

常用脱敏工具库

# 安装：pip install faker
from faker import Faker
import random
import re
fake = Faker('zh_CN')  # 中文环境

基础脱敏函数

class DataMasker:
    """数据脱敏工具类"""
    @staticmethod
    def mask_phone(phone: str) -> str:
        """手机号脱敏：138****1234"""
        if len(phone) != 11:
            return phone
        return phone[:3] + '****' + phone[7:]
    @staticmethod
    def mask_id_card(id_card: str) -> str:
        """身份证脱敏：1101011990****1234"""
        if len(id_card) != 18:
            return id_card
        return id_card[:6] + '********' + id_card[14:]
    @staticmethod
    def mask_name(name: str) -> str:
        """姓名脱敏：张**"""
        if len(name) <= 1:
            return name
        return name[0] + '*' * (len(name) - 1)
    @staticmethod
    def mask_email(email: str) -> str:
        """邮箱脱敏：abc***@example.com"""
        if '@' not in email:
            return email
        name, domain = email.split('@')
        if len(name) <= 3:
            masked_name = name[:1] + '***'
        else:
            masked_name = name[:3] + '***'
        return f"{masked_name}@{domain}"
    @staticmethod
    def mask_address(address: str) -> str:
        """地址脱敏：北京市朝阳区***"""
        if len(address) <= 6:
            return address[:3] + '***'
        return address[:6] + '***'
    @staticmethod
    def mask_bank_card(card_no: str) -> str:
        """银行卡号脱敏：6222 **** **** 1234"""
        if len(card_no) < 4:
            return card_no
        return card_no[:4] + ' **** **** ' + card_no[-4:]

实际应用案例

def demo_basic_masking():
    """基础脱敏演示"""
    masker = DataMasker()
    test_data = {
        '姓名': '张三丰',
        '手机号': '13812345678',
        '身份证': '110101199001011234',
        '邮箱': 'zhangsan@example.com',
        '地址': '北京市朝阳区建国路88号',
        '银行卡': '6222021234567890'
    }
    print("=== 基础脱敏演示 ===")
    print(f"原始姓名: {test_data['姓名']} -> 脱敏后: {masker.mask_name(test_data['姓名'])}")
    print(f"原始手机: {test_data['手机号']} -> 脱敏后: {masker.mask_phone(test_data['手机号'])}")
    print(f"原始身份证: {test_data['身份证']} -> 脱敏后: {masker.mask_id_card(test_data['身份证'])}")
    print(f"原始邮箱: {test_data['邮箱']} -> 脱敏后: {masker.mask_email(test_data['邮箱'])}")
    print(f"原始地址: {test_data['地址']} -> 脱敏后: {masker.mask_address(test_data['地址'])}")
    print(f"原始银行卡: {test_data['银行卡']} -> 脱敏后: {masker.mask_bank_card(test_data['银行卡'])}")
demo_basic_masking()

使用Faker生成假数据替换

def demo_faker_masking():
    """使用Faker生成假数据替换"""
    print("\n=== Faker假数据替换 ===")
    # 原始数据
    original = {
        'name': '李四',
        'phone': '13912345678',
        'email': 'lisi@company.com',
        'company': '某某科技有限公司'
    }
    # 生成替代数据
    masked = {
        'name': fake.name(),
        'phone': fake.phone_number(),
        'email': fake.email(),
        'company': fake.company()
    }
    print("原始数据:", original)
    print("脱敏数据:", masked)
demo_faker_masking()

批量数据脱敏处理

import pandas as pd
def batch_masking_example():
    """批量数据处理"""
    # 创建示例数据
    data = {
        'name': ['张三', '李四', '王五', '赵六'],
        'phone': ['13800138001', '13900139002', '13700137003', '13600136004'],
        'id_card': ['110101199001011111', '110101199002022222', 
                    '110101199003033333', '110101199004044444'],
        'salary': [15000, 20000, 18000, 22000]
    }
    df = pd.DataFrame(data)
    masker = DataMasker()
    print("=== 批量数据脱敏 ===")
    print("原始数据:")
    print(df)
    # 批量脱敏
    df['name'] = df['name'].apply(masker.mask_name)
    df['phone'] = df['phone'].apply(masker.mask_phone)
    df['id_card'] = df['id_card'].apply(masker.mask_id_card)
    # 薪资四舍五入处理
    df['salary'] = df['salary'].apply(lambda x: round(x, -3))  # 保留千位
    print("\n脱敏后数据:")
    print(df)
batch_masking_example()

高级脱敏模式（差分隐私）

import numpy as np
class DifferentialPrivacyMasker:
    """差分隐私脱敏"""
    @staticmethod
    def add_laplace_noise(data, epsilon=1.0, sensitivity=1.0):
        """
        添加拉普拉斯噪声实现差分隐私
        epsilon: 隐私预算（越小隐私保护越强）
        sensitivity: 敏感度
        """
        noise = np.random.laplace(0, sensitivity/epsilon, len(data))
        return [x + n for x, n in zip(data, noise)]
    def mask_numerical_data(self, data, epsilon=0.5):
        """数值数据脱敏"""
        return self.add_laplace_noise(data, epsilon)
def demo_differential_privacy():
    """差分隐私演示"""
    masker = DifferentialPrivacyMasker()
    # 原始薪资数据
    salaries = [15000, 20000, 18000, 22000, 16000]
    print("\n=== 差分隐私脱敏 ===")
    print(f"原始薪资: {salaries}")
    # 不同隐私保护级别
    for eps in [0.1, 0.5, 1.0]:
        masked = masker.mask_numerical_data(salaries, epsilon=eps)
        print(f"隐私预算 ε={eps}: {[round(x) for x in masked]}")
demo_differential_privacy()

正则表达式智能脱敏

class SmartMasker:
    """智能识别并脱敏"""
    def __init__(self):
        self.patterns = {
            'phone': (r'1[3-9]\d{9}', lambda m: m.group()[:3] + '****' + m.group()[7:]),
            'email': (r'\b[A-Za-z0-9._%+-]+@[A-Za-z0-9.-]+\.[A-Z|a-z]{2,}\b', 
                     lambda m: m.group().split('@')[0][:2] + '***@' + m.group().split('@')[1]),
            'id_card': (r'\d{17}[\dXx]', lambda m: m.group()[:6] + '********' + m.group()[14:]),
            'bank_card': (r'\d{16,19}', lambda m: m.group()[:4] + ' **** **** ' + m.group()[-4:])
        }
    def smart_mask(self, text):
        """智能识别并脱敏文本中的敏感信息"""
        result = text
        for pattern_type, (pattern, mask_func) in self.patterns.items():
            result = re.sub(pattern, mask_func, result)
        return result
def demo_smart_masking():
    """智能脱敏演示"""
    masker = SmartMasker()
    text = """
    用户信息：
    姓名：张三
    手机：13812345678
    邮箱：zhangsan@example.com
    身份证：110101199001011234
    银行卡：6222021234567890
    """
    print("=== 智能脱敏 ===")
    print("原始文本:")
    print(text)
    print("\n脱敏后:")
    print(masker.smart_mask(text))
demo_smart_masking()

完整的数据脱敏流程

class DataMaskingPipeline:
    """完整的数据脱敏流水线"""
    def __init__(self, config=None):
        self.config = config or {
            'name': 'mask_name',
            'phone': 'mask_phone',
            'id_card': 'mask_id_card',
            'email': 'mask_email',
            'address': 'mask_address',
            'salary': 'round_to_thousand'
        }
        self.masker = DataMasker()
    def mask_record(self, record):
        """脱敏单条记录"""
        masked = {}
        for field, value in record.items():
            method = self.config.get(field)
            if method and hasattr(self.masker, method):
                masked[field] = getattr(self.masker, method)(value)
            elif method == 'round_to_thousand':
                masked[field] = round(value, -3) if isinstance(value, (int, float)) else value
            else:
                masked[field] = value
        return masked
    def mask_dataset(self, dataset):
        """脱敏整个数据集"""
        return [self.mask_record(record) for record in dataset]
def demo_complete_pipeline():
    """完整流水线演示"""
    pipeline = DataMaskingPipeline()
    # 示例数据
    dataset = [
        {'name': '张三', 'phone': '13812345678', 'id_card': '110101199001011234', 
         'email': 'zhangsan@example.com', 'salary': 15000},
        {'name': '李四', 'phone': '13987654321', 'id_card': '110101199002022345',
         'email': 'lisi@example.com', 'salary': 20000}
    ]
    print("\n=== 完整脱敏流程 ===")
    print("原始数据集:")
    for record in dataset:
        print(record)
    print("\n脱敏后数据集:")
    masked_dataset = pipeline.mask_dataset(dataset)
    for record in masked_dataset:
        print(record)
demo_complete_pipeline()