跳到主要内容

functools

问题​

functools 模块有哪些常用工具？lru_cache、partial、reduce、singledispatch 怎么用？

答案​

lru_cache — 函数缓存​

from functools import lru_cache, cache

@lru_cache(maxsize=128)  # LRU 缓存，最多 128 个结果
def fibonacci(n: int) -> int:
    if n < 2:
        return n
    return fibonacci(n - 1) + fibonacci(n - 2)

fibonacci(100)
print(fibonacci.cache_info())  # CacheInfo(hits=98, misses=101, ...)
fibonacci.cache_clear()        # 清除缓存

# Python 3.9+：无限缓存
@cache  # 等价于 @lru_cache(maxsize=None)
def expensive(x):
    return x ** 2

partial — 偏函数​

from functools import partial

def power(base, exponent):
    return base ** exponent

square = partial(power, exponent=2)
cube = partial(power, exponent=3)

print(square(5))  # 25
print(cube(5))    # 125

# 常用于回调
import json
parse_json = partial(json.loads, encoding="utf-8")

reduce — 归约​

from functools import reduce

# 累积运算
result = reduce(lambda acc, x: acc + x, [1, 2, 3, 4, 5])  # 15
result = reduce(lambda acc, x: acc * x, [1, 2, 3, 4, 5])  # 120

# 带初始值
result = reduce(lambda acc, x: acc + x, [1, 2, 3], 10)     # 16

singledispatch — 单分派泛函数​

from functools import singledispatch

@singledispatch
def process(data):
    raise TypeError(f"不支持的类型: {type(data)}")

@process.register(str)
def _(data: str):
    return data.upper()

@process.register(list)
def _(data: list):
    return [process(item) for item in data]

@process.register(int)
def _(data: int):
    return data * 2

print(process("hello"))    # "HELLO"
print(process(42))         # 84
print(process([1, "a"]))   # [2, "A"]

wraps — 保留元信息​

from functools import wraps

def decorator(func):
    @wraps(func)  # 保留原函数的 __name__、__doc__ 等
    def wrapper(*args, **kwargs):
        return func(*args, **kwargs)
    return wrapper

常见面试问题​

Q1: lru_cache 的参数必须是可哈希的吗？​

答案：

是。lru_cache 用参数的哈希值作为缓存键，所以参数必须是可哈希类型。list、dict 等可变类型不可哈希，不能直接作为被缓存函数的参数。

@lru_cache
def func(lst):     # ❌ TypeError: unhashable type: 'list'
    pass

@lru_cache
def func(t: tuple):  # ✅ tuple 可哈希
    pass

相关链接​