Deep set python dictionary

There is now a way to deep set values dynamically, without changing the container type to a defaultdict or some other dynamically creating type: glom.assign().

A usage example with your case:

import glom

target = {}
path = 'level_one.level_two.test'

glom.assign(target, path, 'hello world', missing=dict)
# {'level_one': {'level_two': {'test': 'hello world'}}}

Notice we passed the missing=dict, telling glom to create missing keys with the built-in dict constructor. You'll also need to pip install glom, but it's pure-Python and compatible with Python 2, 3, and PyPy.

There's a lot more you can do with glom, especially around deep getting and setting. I should know, since (full disclosure) I created it. That means if you find a gap, you should let me know!

We need a defaultdict that will keep making new instances of itself all the way down.

import collections

recursive_dict = lambda: collections.defaultdict(recursive_dict)

mydict = recursive_dict()

At this point, you can simplify your deepGet to just use operator.getitem instead of the lambda you have now. operator.getitem is "shorthand" for lambda a, b: a[b].

As for setting a key, you can just write a loop:

keys = ['level_one', 'level_two', 'test']
d = mydict
for key in keys[:-1]:
    d = d[key]
d[keys[-1]] = 'Hello World'

If dictionaries are missing, the defaultdict will silently create them, no need to check for them first.

Converting this to a reduce() call is left as an exercise for the reader.