这个装饰器的类型标注正确吗?

Lau*_*che -3 python python-3.x mypy

def timer(func: Callable[..., Any]) -> Callable[..., Any]:
"""Calculates the runtime of a function, and outputs it to logging.DEBUG."""

@wraps(func)
def wrapper(*args, **kwargs):
    start = perf_counter()
    value = func(*args, **kwargs)
    end = perf_counter()
    _logger = logging.getLogger(__name__ + '.' + func.__name__)
    _logger.debug(' runtime: {:.4f} seconds'.format(end - start))
    return value

return wrapper
Run Code Online (Sandbox Code Playgroud)

Nat*_*han 6

从 Python 3.10 开始,您可以使用以下方法通过装饰器轻松保留所有类型信息ParamSpec

from collections.abc import Callable
from functools import wraps
from typing import TypeVar, ParamSpec

T = TypeVar('T')
P = ParamSpec('P')

def timer(func: Callable[P, T]) -> Callable[P, T]:
    """Calculates the runtime of a function, and outputs it to logging.DEBUG."""

    @wraps(func)
    def wrapper(*args: P.args, **kwargs: P.kwargs) -> T:
        start = perf_counter()
        value = func(*args, **kwargs)
        end = perf_counter()
        _logger = logging.getLogger(__name__ + "." + func.__name__)
        _logger.debug(" runtime: {:.4f} seconds".format(end - start))
        return value

    return wrapper
Run Code Online (Sandbox Code Playgroud)

(对于 python < 3.10,但更喜欢 juanpa.arrivilillaga 的答案来保留有关输入类型的信息)

这里的缩进似乎有点不对劲,但除此之外,是的,类型并没有不正确。不过,你可以让它更精确一点。您输出的函数与作为输入的函数具有相同的返回类型,但您没有注意到这一点。

特别是,你可以这样说

from collections.abc import Callable
from functools import wraps
from typing import TypeVar, ParamSpec

T = TypeVar('T')
P = ParamSpec('P')

def timer(func: Callable[P, T]) -> Callable[P, T]:
    """Calculates the runtime of a function, and outputs it to logging.DEBUG."""

    @wraps(func)
    def wrapper(*args: P.args, **kwargs: P.kwargs) -> T:
        start = perf_counter()
        value = func(*args, **kwargs)
        end = perf_counter()
        _logger = logging.getLogger(__name__ + "." + func.__name__)
        _logger.debug(" runtime: {:.4f} seconds".format(end - start))
        return value

    return wrapper
Run Code Online (Sandbox Code Playgroud)

看起来你应该能够对 args/kwargs 做同样的事情,但我在自己的打字经验中没有遇到过这种情况,所以我不能确切地说出如何做。编辑 -有关输入这些内容的更多信息,请参阅此 GitHub 问题;这似乎不可能(还?)。

我想你也可以说

from typing import Callable, TypeVar

T = TypeVar("T")
def timer(func: Callable[..., T]) -> Callable[..., T]:
Run Code Online (Sandbox Code Playgroud)

但这似乎没那么有用。


jua*_*aga 5

这种方法的问题在于,现在MyPy丢失了返回类型,或者简而言之,它退化为Any,因此请考虑:

import logging
from typing import Callable, Any
from time import perf_counter
from functools import wraps

def timer(func: Callable[..., Any]) -> Callable[..., Any]:
    """Calculates the runtime of a function, and outputs it to logging.DEBUG."""

    @wraps(func)
    def wrapper(*args, **kwargs):
        start = perf_counter()
        value = func(*args, **kwargs)
        end = perf_counter()
        _logger = logging.getLogger(__name__ + '.' + func.__name__)
        _logger.debug(' runtime: {:.4f} seconds'.format(end - start))
        return value
    return wrapper

@timer
def func(x:int) -> int:
    return x


def string_func(s: str):
    return s[:]

x = 42 * func(42)

reveal_type(x)

string_func(x)
Run Code Online (Sandbox Code Playgroud)

使用方法:

(py37) Juans-MacBook-Pro:tempdata juan$ mypy --version
mypy 0.641
Run Code Online (Sandbox Code Playgroud)

如果我尝试对此进行类型检查,它会通过!

(py37) Juans-MacBook-Pro:tempdata juan$ mypy typing_decorators.py
typing_decorators.py:29: error: Revealed type is 'Any'
Run Code Online (Sandbox Code Playgroud)

我想找到一种解决方案,其中要保留精确的参数,即使用a TypeVarcast包装器,这样MyPy可以确切地知道类型(假设原始函数被标注):

import logging
from typing import Callable, Any, TypeVar, cast
from time import perf_counter
from functools import wraps


F = TypeVar('F', bound=Callable[..., Any])

def timer(func: F) -> F:
    """Calculates the runtime of a function, and outputs it to logging.DEBUG."""

    @wraps(func)
    def wrapper(*args, **kwargs):
        start = perf_counter()
        value = func(*args, **kwargs)
        end = perf_counter()
        _logger = logging.getLogger(__name__ + '.' + func.__name__)
        _logger.debug(' runtime: {:.4f} seconds'.format(end - start))
        return value
    return cast(F, wrapper)

@timer
def func(x:int) -> int:
    return x


def string_func(s: str):
    return s[:]

x = 42 * func(42)

reveal_type(x)

string_func(x)
Run Code Online (Sandbox Code Playgroud)

在这种情况下:

(py37) Juans-MacBook-Pro:tempdata juan$ mypy typing_decorators.py
typing_decorators.py:32: error: Revealed type is 'builtins.int'
typing_decorators.py:34: error: Argument 1 to "string_func" has incompatible type "int"; expected "str"
Run Code Online (Sandbox Code Playgroud)