""" Day 047 — 描述符（Descriptor）进阶用法与避坑 本示例展示： 1. 数据描述符 vs 非数据描述符的优先级差异 2. 描述符的常见陷阱及解决方案 3. 实际工程中的描述符模式 """ # ============================================================ # 1. 数据描述符 vs 非数据描述符：优先级差异 # ============================================================ class DataDescriptor: """数据描述符：同时实现 __get__ 和 __set__""" def __get__(self, obj, objtype=None): if obj is None: return "类访问 → DataDescriptor" return f"数据描述符拦截 → {getattr(obj, '_val', '无值')}" def __set__(self, obj, value): print(f" ⚡ 数据描述符 __set__ 触发 (value={value!r})") obj._val = value class NonDataDescriptor: """非数据描述符：只实现 __get__""" def __get__(self, obj, objtype=None): if obj is None: return "类访问 → NonDataDescriptor" return f"非数据描述符 → {getattr(obj, '_val2', '无值')}" class PriorityTest: data_desc = DataDescriptor() non_data_desc = NonDataDescriptor() print("=" * 60) print("1. 数据描述符 vs 非数据描述符的优先级") print("=" * 60) obj = PriorityTest() # 数据描述符：实例 __dict__ 无法覆盖 print("\n数据描述符测试:") obj.data_desc = "尝试通过实例覆盖" print(f" obj.data_desc = {obj.data_desc}") # 数据描述符仍然拦截 print(f" obj.__dict__ 有 _val 吗？{'_val' in obj.__dict__}") # True print(f" obj.__dict__['_val'] = {obj.__dict__.get('_val')}") # "尝试通过实例覆盖" # 非数据描述符：实例 __dict__ 可以覆盖 print("\n非数据描述符测试:") obj.non_data_desc = "实例覆盖了描述符" print(f" obj.non_data_desc = {obj.non_data_desc}") # 实例值胜出！ print(f" 但类访问仍触发描述符:") print(f" PriorityTest.non_data_desc → {PriorityTest.non_data_desc}") # 清理 del obj.non_data_desc print(f" del 后: obj.non_data_desc = {obj.non_data_desc}") # 描述符回来 # ============================================================ # 2. 陷阱：方法被实例属性覆盖 # ============================================================ print("\n" + "=" * 60) print("2. 陷阱：方法被实例属性覆盖") print("=" * 60) class Dog: def __init__(self, name): self.name = name def bark(self): return f"{self.name} says: Woof!" dog = Dog("Buddy") print(f"\n正常调用: {dog.bark()}") # ⚠️ 危险操作：覆盖方法！ dog.bark = "覆盖了方法" print(f"dog.bark = {dog.bark}") # "覆盖了方法" try: dog.bark() # TypeError! except TypeError as e: print(f"❌ 调用报错: {e}") # 解决方案：用描述符保护方法 class ProtectedMethod: """防止方法被实例属性覆盖的描述符""" def __init__(self, func): self.func = func self.__doc__ = func.__doc__ self.__name__ = func.__name__ def __set_name__(self, owner, name): self.name = name def __get__(self, obj, objtype=None): if obj is None: return self.func # 类访问返回原始函数 # 创建绑定方法 import types return types.MethodType(self.func, obj) def __set__(self, obj, value): raise AttributeError(f"不能覆盖受保护的方法 '{self.name}'") class SafeDog: def __init__(self, name): self.name = name @ProtectedMethod def bark(self): return f"{self.name} says: Woof!" safe_dog = SafeDog("Rex") print(f"\n受保护的方法: {safe_dog.bark()}") try: safe_dog.bark = "尝试覆盖" except AttributeError as e: print(f"✅ 拦截成功: {e}") # ============================================================ # 3. 陷阱：描述符实例共享状态 # ============================================================ print("\n" + "=" * 60) print("3. 陷阱：描述符实例共享状态") print("=" * 60) # ❌ 错误示范 class SharedCounter: """所有实例共享同一个计数器！""" def __get__(self, obj, objtype=None): return self.count if hasattr(self, 'count') else 0 def __set__(self, obj, value): self.count = value # 存到描述符实例上，所有对象共享 class WrongClass: counter = SharedCounter() a = WrongClass() b = WrongClass() a.counter = 10 print(f"\n❌ 共享状态错误:") print(f" a.counter = {a.counter}") # 10 print(f" b.counter = {b.counter}") # 也是 10！共享了 # ✅ 正确示范 class IsolatedCounter: """每个实例独立的计数器""" def __set_name__(self, owner, name): self.name = name def __get__(self, obj, objtype=None): if obj is None: return self return obj.__dict__.get(self.name, 0) def __set__(self, obj, value): obj.__dict__[self.name] = value class RightClass: counter = IsolatedCounter() a2 = RightClass() b2 = RightClass() a2.counter = 10 print(f"\n✅ 独立状态正确:") print(f" a2.counter = {a2.counter}") # 10 print(f" b2.counter = {b2.counter}") # 0，独立 # ============================================================ # 4. 实际应用：只读属性描述符 # ============================================================ print("\n" + "=" * 60) print("4. 只读属性描述符") print("=" * 60) class ReadOnly: """只读属性：首次设置后不可修改""" def __set_name__(self, owner, name): self.name = name self.private_name = f"_{name}" def __get__(self, obj, objtype=None): if obj is None: return self if not hasattr(obj, self.private_name): raise AttributeError(f"{self.name} 尚未设置") return getattr(obj, self.private_name) def __set__(self, obj, value): if hasattr(obj, self.private_name): raise AttributeError(f"{self.name} 是只读属性，不可修改") setattr(obj, self.private_name, value) class ImmutableRecord: id = ReadOnly() created_at = ReadOnly() def __init__(self, id, created_at): self.id = id self.created_at = created_at def __repr__(self): return f"ImmutableRecord(id={self.id!r}, created_at={self.created_at!r})" record = ImmutableRecord("REC-001", "2026-07-03") print(f"\n{record}") try: record.id = "MODIFIED" # ❌ 只读！ except AttributeError as e: print(f"✅ 拦截成功: {e}") # ============================================================ # 5. 实际应用：带日志的描述符 # ============================================================ print("\n" + "=" * 60) print("5. 带日志的描述符") print("=" * 60) class LoggedField: """自动记录所有属性操作的描述符""" def __init__(self, field_type=str): self.field_type = field_type self.logs = [] def __set_name__(self, owner, name): self.name = name self.private_name = f"_{name}" def __get__(self, obj, objtype=None): if obj is None: return self value = getattr(obj, self.private_name, None) log = f"[GET] {self.name} = {value!r}" self.logs.append(log) print(f" 📋 {log}") return value def __set__(self, obj, value): value = self.field_type(value) # 强制类型转换 old = getattr(obj, self.private_name, None) setattr(obj, self.private_name, value) log = f"[SET] {self.name}: {old!r} → {value!r}" self.logs.append(log) print(f" 📋 {log}") def __delete__(self, obj): old = getattr(obj, self.private_name, None) delattr(obj, self.private_name) log = f"[DEL] {self.name}: {old!r} → None" self.logs.append(log) print(f" 📋 {log}") def get_logs(self): return list(self.logs) class Config: host = LoggedField(str) port = LoggedField(int) debug = LoggedField(bool) config = Config() print("\n设置配置:") config.host = "localhost" config.port = "8080" # 自动转为 int config.debug = "yes" # 自动转为 bool (True) print(f"\n host={config.host}, port={config.port}, debug={config.debug}") print(f"\n 📊 操作日志: {Config.host.get_logs()}") print("\n✅ 所有测试通过！")