Cache在Python中的实现-AOP内置实现

最新推荐文章于 2025-12-29 13:38:51 发布

原创最新推荐文章于 2025-12-29 13:38:51 发布 · 1k 阅读

0 ·

CC 4.0 BY-SA版权

文章标签：

#AOP #PYthon #Cache

Python 同时被 2 个专栏收录

4 篇文章

订阅专栏

AOP

1 篇文章

订阅专栏

本文介绍Python中利用functools.lru_cache装饰器实现的LRU缓存机制。通过示例展示了如何设置缓存大小及缓存溢出的情况。当缓存未满时，相同参数的函数调用会返回相同的输出；而当缓存溢出时，即使参数相同也会因重新计算而得到不同的结果。

Python自带的AOP Cache

JAVA中可以通过AOP很容易的实现方法的cache, 其实Python比JAVA做的更进一步,直接在核心库中就实现了.

示例

from functools import lru_cache
import time
import numpy as np

#maxsiez设置了这个方法的case size只有2个,超过就会删除最近最少使用
@lru_cache(maxsize=2)
def test_lru(seed):   
    return np.random.randint(9)

Cache不溢出

for i in range(0, 6):
    print(f'2 diff paras, not over maxsize. input:{i%2} output:{test_lru(i%2)}')

##没有超出Cache的size(2),所以反复调用,相同参数返回相同值

2 diff paras, not over maxsize. input:0 output:8
2 diff paras, not over maxsize. input:1 output:3
2 diff paras, not over maxsize. input:0 output:8
2 diff paras, not over maxsize. input:1 output:3
2 diff paras, not over maxsize. input:0 output:8
2 diff paras, not over maxsize. input:1 output:3

Cache溢出

for i in range(0, 6):
    print(f'3 diff paras, over maxsize. input:{i%3},  output:{test_lru(i%3)}')

##超出Cache的size(2),有3个不同的参数,当再次调用相同参数时,得到了不同的结果(也就是再次调用时,不是从Cache中取到的)

3 diff paras, over maxsize. input:0,  output:8
3 diff paras, over maxsize. input:1,  output:3
3 diff paras, over maxsize. input:2,  output:7
3 diff paras, over maxsize. input:0,  output:4
3 diff paras, over maxsize. input:1,  output:3
3 diff paras, over maxsize. input:2,  output:1