Description

Nested dictionary structures emerge every time there is some sort hierarchy in the data. Trees, archives, the chapters and sections in a book, these are all examples where you will likely find nested loops of data.

Python does not have a built-in data type for nested mappings. Dictionaries can be used, however there many inconveniences and limitations. Two above all: getting items requires to open and close several square brackets (eg. d[level1][level2]...[levelN]), and iterating will only access the first layer, so nested for loops are needed to iterate through all the values.

The ndicts package aims to tackle the main issues of nested dictionaries, exposing an interface with minimum differences from dictionaries themselves. NestedDict is a MutableMapping at its core, therefore all familiar dictionary methods are available and the overall behaviour similar.

If you need to perform simple mathematical operations with your nested data, use DataDict. In addition to allowing arithmetics, DataDicts borrow some methods that you would expect from a pandas DataFrame.

Finally, this is a simple project for simple needs. Consider using pandas MultiIndex for more functionalities!

Installation

Install ndicts with pip.

pip install ndicts

Overview

Import NestedDict and DataDict.

>>> from ndicts import DataDict, NestedDict

Create a NestedDict from a dictionary.

>>> book = {
...     "Book 1": {
...         "Section 1": "The Eve of the War",
...         "Section 2": "The Falling Star"
...     },
...     "Book 2": {
...         "Section 1": "Under Foot", 
...         "Section 2": {"Paragraph 1": "After eating we crept back to the scullery"}
...     }
... }
>>> nd = NestedDict(book)

Get items more conveniently than with standard dictionaries.

>>> # NestedDict
>>> nd["Book 1", "Section 1"] 
'The Eve of the War'
>>> # dict
>>> book["Book 1"]["Section 2"]
'The Falling Star'

Iterate over a NestedDict.

>>> for key in nd:
...     print(key)
('Book 1', 'Section 1')
('Book 1', 'Section 2')
('Book 2', 'Section 1')
('Book 2', 'Section 2', 'Paragraph 1')

Documentation

https://edd313.github.io/ndicts/

Licence

ndicts is licensed under the MIT license.