This is how it is written in Python 2. Come to Python version 3.6:

def dict_deep_merge(a, b): if not b: return a return {**a, **b, **{k: dict_deep_merge(a[k], b[k]) for k in set(a) & set(b)}}

It should be more efficient. Everything else is pretty much the same.

It’s not a god-level code, nor is it difficult to understand. Just recursively merge values with the same key. What you need to know:

• dict’s items method

Addition of

• tuples

• Gathering together

• The meaning of dict parameters

函数的作用合并两个dict 比如 a = {'a': {'A': 1}, 'b': 1} b = {'a': {'B': 1}} 合并成 {'a': {'A': 1, 'B': 1}, 'b': 1} set(b) & set(a)是取a,c的key交集，如上a,b的key交集为a, 再递归运行子dict

Ask a question, is there something wrong with the code? If the value in the same key is a string, will the items function report an error?