The following code works but it's very ungainly:

        a = 0
        for k in keys:
            a = a   1
            if a == 1:
                k1 = k
            if a == 2:
                k2 = k
            if a == 3:
                k3 = k
        if a == 1:
            list_of_dict = sorted(list_of_dict, key = lambda ele: ele[k1], reverse = rev)
        if a == 2:
            list_of_dict = sorted(list_of_dict, key = lambda ele: ele[k1][k2], reverse = rev)
        if a == 3:
            list_of_dict = sorted(list_of_dict, key = lambda ele: ele[k1][k2][k3], reverse = rev)

My question is if there is any way to express this more succinctly. (note, reverse references the [unshown] variable rev, so I figured there must be a way to do the same with ele)

CodePudding user response：

This can be implemented with closures:

def nested_itemgetter(keys):
    def getter(val):
        for key in keys:
            val = val[key]
        return val
    return getter

Or class:

class NestedItemGetter:
    def __init__(self, keys):
        self.keys = keys

    def __call__(self, val):
        for key in self.keys:
            val = val[key]
        return val

Usage:

list_of_dict = sorted(list_of_dict, key=nested_itemgetter(keys), reverse=rev)
list_of_dict = sorted(list_of_dict, key=NestedItemGetter(keys), reverse=rev)

CodePudding user response：

Try:

from functools import partial


def nested_itemgetter(ks, d):
    res = d
    for key in ks:
        res = res[key]
    return res


rev = True
list_of_dict = [{"a": {"b": {"c": 1}}}, {"a": {"b": {"c": 2}}}, {"a": {"b": {"c": 3}}}]

keys = ["a", "b", "c"]
res = sorted(list_of_dict, key=partial(nested_itemgetter, keys), reverse=rev)

print(res)

Output

[{'a': {'b': {'c': 3}}}, {'a': {'b': {'c': 2}}}, {'a': {'b': {'c': 1}}}]