umapper

A small Python utility for wrangling dictionary keys: translate between snake_case, camelCase and PascalCase, turn nested dicts into attribute-accessible objects, and merge multiple dicts in one shot.

Installation

pip install umapper

Features

• Case translation — recursively remap all keys in a dict (or list of dicts) to snake, camel, or pascal case
• Object conversion — turn nested dicts into plain objects so you can write response.userId instead of response["userId"]
• Dict assembly — merge several dicts at once, translate their keys, strip None values, and inject extra fields, all in one call

Usage

Case translation

from umapper import Case, translate_case

data = {
    'OutterField': {'inner_field': 123, 'sibling_inner_field': 321},
    'siblingField': 'Hello',
}

translate_case(data, Case.SNAKE)
# {'outter_field': {'inner_field': 123, 'sibling_inner_field': 321}, 'sibling_field': 'Hello'}

translate_case(data, Case.CAMEL)
# {'outterField': {'innerField': 123, 'siblingInnerField': 321}, 'siblingField': 'Hello'}

translate_case(data, Case.PASCAL)
# {'OutterField': {'InnerField': 123, 'SiblingInnerField': 321}, 'SiblingField': 'Hello'}

Translation works on lists too — each element is converted independently.

Translation works on tuples too. Scalar values (strings, numbers, etc.) are returned as-is.

Object conversion

from umapper import Case, translate_case, convert_to_object

data = {'userId': 42, 'homeAddress': {'city': 'São Paulo'}}
obj = convert_to_object(translate_case(data, Case.SNAKE))

print(obj.user_id)             # 42
print(obj.home_address.city)   # São Paulo

translate_case is optional — use it beforehand when you want a specific attribute naming style on the resulting object.

Assembling dicts

assemble_dicts merges any number of positional dicts, translates all keys to a target case (camelCase by default), and lets you inject extra fields as keyword arguments.

from umapper import assemble_dicts

user   = {'first_name': 'Ada', 'last_name': 'Lovelace'}
coords = {'lat': -23.5, 'lng': -46.6}

result = assemble_dicts(user, coords, active=True)
# {
#   'firstName': 'Ada',
#   'lastName': 'Lovelace',
#   'lat': -23.5,
#   'lng': -46.6,
#   'active': True,
# }

Options:

# Strip None values and use snake_case output
result = assemble_dicts(user, coords, mapping_case=Case.SNAKE, include_nones=False)

Registering custom mapping classes

Some libraries provide dict-like objects that don't inherit from collections.abc.Mapping. Register them so umapper traverses them correctly:

from umapper import register_mapping_class
register_mapping_class(MyCustomDictLike)

API reference

Full API docs are available on Read the Docs.