class User:
    def __init__(self, name, email, temp_token):
        self.name = name
        self.email = email
        self.temp_token = temp_token

user = User("Alice", "alice@example.com", "x192838")

# 基礎(chǔ)刪除操作
delattr(user, 'temp_token')
print(hasattr(user, 'temp_token'))  # 輸出: False

class UserResponse:
    def __init__(self, uid, name, password_hash, ssn, salary):
        self.uid = uid
        self.name = name
        self.password_hash = password_hash
        self.ssn = ssn  # 社保號
        self.salary = salary # 薪資

    def get_cleaned_data(self, role):
        """
        根據(jù)角色動態(tài)清理屬性
        """
        # 定義不同角色的權(quán)限映射
        restrictions = {
            'guest': ['name', 'password_hash', 'ssn', 'salary'],
            'user': ['password_hash', 'ssn', 'salary'],
            'admin': [] # 管理員無限制
        }
        
        # 獲取需要刪除的屬性列表
        # 如果角色不存在，默認(rèn)限制所有敏感信息
        fields_to_remove = restrictions.get(role, ['password_hash', 'ssn', 'salary'])
        
        # 核心邏輯：利用 delattr 進(jìn)行動態(tài)刪除
        for field in fields_to_remove:
            # 必須先檢查屬性是否存在，防止報(bào)錯
            if hasattr(self, field):
                delattr(self, field)
        
        # 轉(zhuǎn)換為字典返回
        return self.__dict__

# 測試
user = UserResponse(1, "Bob", "hashed_123", "123-45-6789", 50000)

print("Guest View:", user.get_cleaned_data('guest'))
# 重置對象（因?yàn)樵瓕ο髮傩砸驯粍h）
user = UserResponse(1, "Bob", "hashed_123", "123-45-6789", 50000)
print("Admin View:", user.get_cleaned_data('admin'))

class DataProcessor:
    def __init__(self, raw_data):
        self.raw_data = raw_data  # 假設(shè)這是一個占用 1GB 的巨大列表
        self.is_processed = False

    def extract_features(self):
        print("開始特征提取...")
        # 模擬特征提取邏輯
        self.features = [x * 2 for x in self.raw_data]
        self.is_processed = True
        print("特征提取完成。")

    def cleanup(self):
        """
        條件判斷：僅在處理完成后，且不再需要原始數(shù)據(jù)時進(jìn)行清理
        """
        if self.is_processed and hasattr(self, 'raw_data'):
            print("正在釋放原始數(shù)據(jù)內(nèi)存...")
            delattr(self, 'raw_data')
        elif not self.is_processed:
            print("警告：尚未處理完成，不能清理原始數(shù)據(jù)。")
        else:
            print("原始數(shù)據(jù)已清理或不存在。")

# 模擬使用
processor = DataProcessor(list(range(1000000))) # 模擬大對象
processor.extract_features()
processor.cleanup() # 成功清理
processor.cleanup() # 重復(fù)清理提示

def create_model_class(class_name, allow_delete=False):
    # 基礎(chǔ)屬性
    attrs = {
        'id': None,
        'save': lambda self: print(f"保存 {class_name} 實(shí)例..."),
        '__init__': lambda self, id: setattr(self, 'id', id)
    }

    # 條件判斷：動態(tài)添加或移除 delete 方法
    if allow_delete:
        attrs['delete'] = lambda self: print(f"刪除 {class_name} 實(shí)例...")
    
    # 使用 type 動態(tài)創(chuàng)建類
    ModelClass = type(class_name, (object,), attrs)
    
    # 如果不允許刪除，我們甚至可以從類定義中物理移除該方法（雖然上面的邏輯已經(jīng)不添加了）
    # 但假設(shè)我們是先定義好了一個通用類，現(xiàn)在需要根據(jù)參數(shù)“閹割”它：
    if not allow_delete:
        # 這里的 delattr 作用于類本身，而不是實(shí)例
        if hasattr(ModelClass, 'delete'):
            delattr(ModelClass, 'delete')
            
    return ModelClass

# 創(chuàng)建一個不允許刪除的模型
ReadOnlyUser = create_model_class('ReadOnlyUser', allow_delete=False)
user_instance = ReadOnlyUser(1)
user_instance.save()
try:
    user_instance.delete()
except AttributeError as e:
    print(f"錯誤: {e}") # 報(bào)錯，因?yàn)榉椒ú淮嬖?

print("-" * 20)

# 創(chuàng)建一個允許刪除的模型
FullAccessUser = create_model_class('FullAccessUser', allow_delete=True)
admin_instance = FullAccessUser(99)
admin_instance.delete()

# 不推薦
try:
    delattr(obj, 'maybe_exists')
except AttributeError:
    pass

# 推薦
if hasattr(obj, 'maybe_exists'):
    delattr(obj, 'maybe_exists')

obj = "Hello"
# delattr(obj, '__class__') # 這可能會導(dǎo)致解釋器崩潰或嚴(yán)重錯誤

class SlotClass:
    __slots__ = ['a']

s = SlotClass()
s.a = 10
delattr(s, 'a') # 正常工作
# print(s.a) # 拋出 AttributeError
s.a = 20 # 依然可以重新賦值