Library that allows deep extraction of layered data structures (like JSON).

## Magic Dot

Library that allows deep extraction of layered data structures (like JSON).

### Introduction

Magic Dot encapsulates data to allow the versatile extraction of its contents. It is easier to use than setdefault or try: except that typical extraction from a structured like JSON. Consider the following simplified JSON snipppet curl https://api.github.com/events:

import json
[
{
"type": "PushEvent",
"commits": [
{
"author": {
"name": "Bubba"
}}]}}]
""")

magic_dot can retrieve the first name of the first commit, with a default of “nobody” if any part of that chain is missing with the the following:

from magic_dot import MagicDot, NOT_FOUND
name = MagicDot(data)[0].payload.commits[0].author.name.get("nobody")

Since the incoming JSON can’t be trusted, without magic_dot, you have to verify that each layer is there. This can be done with a try: except, nearly as efficiently, but it is more verbose.

try:
except (IndexError, KeyError):
name = "nobody"

Other features, like pluck, selective exceptions, attribute support, and iteration can lead to cleaner code.

### Features

#### Forgiving NOT_FOUND Handling

Manipulations of the MagicDot structure will raise no exceptions when one of the attributes or keys are not found. Instead it delays this until the get() call that extracts the data at the end. When the get() is called, there are three ways of handling missing data:

Default is to return magic_dot.NOT_FOUND

>>> md.nonexistent.get()
magic_dot.NOT_FOUND

You can request a default value for magic_dot.NOT_FOUND

>>> md.nonexistent.get('bubba')
'bubba'

Or you can enable exceptions when referencing the nonexistent data

>>> md.exception().nonexistent
---------------------------------------------------------------------------
NotFound                                  Traceback (most recent call last)

Exceptions are not enabled by default. They can be enabled during creation I.E MagicDot(data, exception=True) and switched on and off with the MagicDot::exception(exception=False) method.

#### Dict and List Item Handling

When a md[item] is encountered, data will be extracted as follows:

1. If md.__data[item] exists, that is used.

2. If md.__data.item attribute exists it is used.

3. If .exception() is enabled, a NotFound exception is raised.

4. Otherwise md.NOT_FOUND is assigned to the resulting encapsulated data.

#### Attribute Handling

When a md.key is supplied data will be extracted as follows:

1. If md.__data.key attribute exists it is used.

2. If md.__data[key] item exists, it is used.

3. If .exception() is enabled, a NotFound exception is raised.

4. Otherwise md.NOT_FOUND is assigned to the resulting md.__data.

#### Iteration Support

If the currently encapsulated data is an iterable, MagicDot supports iterating over the contained data with the resulting iteration being a MagicDot wrapper around the iterated data.

>>> from collections import namedtuple
>>> data = [1, {'x': 2}, namedtuple('x', 'x')(3)]
>>> for md in MagicDot(data):
...   print(md.get())
1
{'x': 2}
x(x=3)


By default, if an attempt is made to iterate over NOT_FOUND data, a TypeError will be raised. The iteration code can be changed to instead return an empty list. ::

>> md = MagicDot(1, iter_nf_as_empty=True)
>> for x in md.nonexistent:
..   print(md.get())
(prints nothing)

#### Other Operators

Currently, there is one additional operator, MagicDot::pluck(), which if the encapsulated data is a list, it will attempt to extract a named attribute or key from the entire list. The returned value is a MagicDot with the new plucked list.

### Future Enhancement

Future enhancements will be to support many of the Underscore js array and collection capabilities like compact, reject, and count.

## History

### 0.2.0 (2020-04-30)

• This contains some modifications that are not backwards compatible.

• Remove .lists() support.

• Change exception() to be raised when NOT_FOUND is generated instead of at the .get()

• Make documentation changes to reflect the above.

### 0.1.1 (2020-03-19)

• No significant chagnes. Testing github release automations.

### 0.1.0 (2020-03-19)

• First release on PyPI.

## Project details

Uploaded source
Uploaded py2 py3