如何过滤某个嵌套级别的嵌套字典？

Question

我想过滤嵌套字典一定水平，例如：

d = {'fruit': {'sweet': {'green': 'apple', 'red': 'tomato'}, 'bitter': {'green': 'cucumber'}}} 

我想写一个函数filter_at_level(d, 2, lambda r: r == 'green')，这将只返回以下而不修改原始数据：

{'fruit': {'sweet': {'green': 'apple'}, 'bitter': {'green': 'cucumber'}}} 

我应该避免递归吗？它可以有效地写入而无需递归？

Answer 1

from itertools import ifilter 

def f(d, func, l): 
    if l == 0: 
     return {k: d[k] for k in ifilter(func, d)} 
    return {k: f(v, func, l - 1) for k, v in d.iteritems()} 

d = {'fruit': {'sweet': {'green': 'apple', 'red': 'tomato'}, 'bitter': {'green': 'cucumber'}}} 

print f(d, lambda k: k == 'green', 2) 

，或者如果你不希望空元素：

from itertools import ifilter 

def f(d, func, l): 
    if l == 0: 
     return {k: d[k] for k in ifilter(func, d)} 
    tmp = {k: f(v, func, l - 1) for k, v in d.iteritems()} 
    return {k: v for k, v in tmp.iteritems() if v} 

d = {'fruit': {'sweet': {'green': 'apple', 'red': 'tomato'}, 'bitter': {'green': 'cucumber'}, 'x': {'a': 'z'}}} 

print f(d, lambda k: k == 'green', 2)