Convert all keys in a nested dictionary from camelCase to snake_case

Question

I have a dictionary something like this:

{
     'firstName': 'abc',
     'lastName': 'xyz',
     'favoriteMovies': ['Star Wars', 'The lone ranger'],
     'favoriteCountries': [
          {'country': 'China', 'capitalCity': 'Beiging'},
          {'country': 'India', 'capitalCity': 'New Delhi'}
     ]
}

I want to convert it to snake_case like the following

{
    'first_name': 'abc',
    'last_name': 'xyz',
    'favorite_movies': ['Star Wars', 'The lone ranger'],
    'favorite_countries': [
        {'country': 'China', 'capital_city': 'Beiging'},
        {'country': 'India', 'capital_city': 'New Delhi'}
     ]
}

The dictionary may be of any length depth.

My current solution is

import re

def convert_snake_case_to_camel_case(data):
    required_dict = {}

    for key, value in data.items():
        if type(value) == str:
            new_key = re.sub("([a-z0-9])([A-Z])", r"\1_\2", key).lower()
            required_dict[new_key] = value
        elif type(value) == list and all(list(map(lambda _: isinstance(_, str), value))):
            new_key = re.sub("([a-z0-9])([A-Z])", r"\1_\2", key).lower()
            required_dict[new_key] = value
        elif type(value) == list and all(list(map(lambda _: isinstance(_, dict), value))):
            new_key = re.sub("([a-z0-9])([A-Z])", r"\1_\2", key).lower()
            required_dict[new_key] = list(filter(convert_snake_case_to_camel_case, value))
    return required_dict

But I m not getting the expected result for the nested data.

What have you tried so far? Can you share your code snippet? Please go through the stackoverflow guidelines on how to ask a questions, if you are unaware of it. Link : http://stackoverflow.com/help/minimal-reproducible-example — High-Octane, Feb 10 '20 at 10:20
Based on your question it looks like you want to achieve the opposite of what you say.. upper one is camelCase lower one is snake_case. — Michael K, Feb 10 '20 at 10:23
@High-Octane, I have done it manually but feels wrong since the data i m working on is dynamic — udeep shrestha, Feb 10 '20 at 10:26
My bad. I was using filter function while i should be using map function. Thank you guys — udeep shrestha, Feb 10 '20 at 11:37

score 9 · Answer 1 · answered Feb 10 '20 at 16:52

Short recursive version:

import re
def to_snake(s):
  return re.sub('([A-Z]\w+$)', '_\\1', s).lower()

def t_dict(d):
   if isinstance(d, list):
      return [t_dict(i) if isinstance(i, (dict, list)) else i for i in d]
   return {to_snake(a):t_dict(b) if isinstance(b, (dict, list)) else b for a, b in d.items()}

data = {'firstName': 'abc', 'lastName': 'xyz', 'favoriteMovies': ['Star Wars', 'The lone ranger'], 'favoriteCountries': [{'country': 'China', 'capitalCity': 'Beiging'}, {'country': 'India', 'capitalCity': 'New Delhi'}]} 
print(t_dict(data))

Output:

{'first_name': 'abc', 'last_name': 'xyz', 
 'favorite_movies': ['Star Wars', 'The lone ranger'], 
 'favorite_countries': [
    {'country': 'China', 'capital_city': 'Beiging'}, 
    {'country': 'India', 'capital_city': 'New Delhi'}
  ]
}

This should be the selected answer, however I am using the following regex pattern instead `re.compile(r'(?<!^)(?=[A-Z])')`. — Jenobi, Mar 02 '23 at 17:34

score 5 · Answer 2 · answered Feb 10 '20 at 10:26

5

You could use regex for this.

def camel_to_snake(str):
   return re.sub(r'(?<!^)(?=[A-Z])', '_', str).lower()

Then build another recursive function which converts all the dictionary keys by using the above function.

answered Feb 10 '20 at 10:26

Mihai Alexandru-Ionut

47,092
13
101
128

Phillip Kigenyi · Answer 3 · 2023-02-02T18:01:30.660

3

Use the humps library

Example

import humps
humps.camelize('jack_in_the_box')  # jackInTheBox
humps.decamelize('rubyTuesdays')  # ruby_tuesdays
humps.pascalize('red_robin')  # RedRobin

array = [{"attrOne": "foo"}, {"attrOne": "bar"}]
humps.decamelize(array) # [{"attr_one": "foo"}, {"attr_one": "bar"}]

edited Feb 02 '23 at 18:01

answered Apr 30 '22 at 00:26

Phillip Kigenyi

1,359
14
21

Zaraki Kenpachi · Accepted Answer · 2020-02-10T10:43:22.127

data = {
     'firstName': 'abc',
     'lastName': 'xyz',
     'favoriteMovies': ['Star Wars', 'The lone ranger'],
     'favoriteCountries': [
          {'country': 'China', 'capitalCity': 'Beiging'},
          {'country': 'India', 'capitalCity': 'New Delhi'}
     ]
}

new_data = {}
for key, value in data.items():
    new_key_list = ['_' + x.lower() if x.isupper() else x for x in key]
    new_key = ''.join(new_key_list)
    if isinstance(value[0],dict):
        new_value = []
        for item in value:
            temp_dict = {}
            for key2, value2 in item.items():
                new_key_list = ['_' + x.lower() if x.isupper() else x for x in key2]
                new_key = ''.join(new_key_list)
                temp_dict[new_key] = value2
            new_value.append(temp_dict)
        new_data[new_key] = new_value
    else:
        new_data[new_key] = value

Output:

{'first_name': 'abc', 'last_name': 'xyz', 'favorite_movies': ['Star Wars', 'The lone ranger'], 'capital_city': [{'country': 'China', 'capital_city': 'Beiging'}, {'country': 'India', 'capital_city': 'New Delhi'}]}

If our dictionary is in a different structure, and includes integers / floats this solution does not work. It may work for the particular dictionary, but errors if key values are ints. Solution by @Ajax1234 works for any pattern. — Jenobi, Mar 02 '23 at 17:38

score 0 · Answer 5 · answered Jan 01 '23 at 21:05

Here's the working code

def convert_case(str_camelcase):
    # This function takes in a string in camelCase and converts it to snake_case
    str_snake_case = ""
    for ele in list(str_camelcase):
        if ele.islower():
            str_snake_case = str_snake_case + ele
        else:
            str_snake_case = str_snake_case + "_" + ele.lower()
    return str_snake_case
    

def convert_json(json_dict):
    # This function takes in a dictionary and converts the keys of the dictionary to snake_case
    temp = {}
    for item in json_dict:
        new_item = convert_case(item)
        temp[new_item]=json_dict[item]
        new_list = []
        if type(json_dict[item]) is list:
            for ele in json_dict[item]:
                if type(ele) is dict:
                    new_list.append(convert_json(ele))
            if len(new_list)!=0:
                temp[new_item] = new_list
        if type(json_dict[item]) is dict:
           # if the value is a dictionary, recursively convert it to snake_case
           temp[new_item] = convert_json(json_dict[item])
    return temp

json = {
   "firstName":"abc",
   "lastName":"xyz",
   "favoriteMovies":[
      "Star Wars",
      "The lone ranger"
   ],
   "favoriteCountries":[
      {
         "country":"China",
         "capitalCity":"Beiging"
      },
      {
         "country":"India",
         "capitalCity":"New Delhi"
      }
   ]
}

print(convert_json(json))

Output:

{
   "first_name":"abc",
   "last_name":"xyz",
   "favorite_movies":[
      "Star Wars",
      "The lone ranger"
   ],
   "favorite_countries":[
      {
         "country":"China",
         "capital_city":"Beiging"
      },
      {
         "country":"India",
         "capital_city":"New Delhi"
      }
   ]
}

Convert all keys in a nested dictionary from camelCase to snake_case

5 Answers5

Linked