SMART DATA

SMART DATA is a package to make your life easier. For example you want to test many JSON (or any other) results from some API in Python. There are plenty of great Python packages to find differences between two complex structres but sometimes you need to omit some of their parts. This package will help you to write this kind of code easier.

Requirements

No package dependecy.

Installation

pip install smart_data

Functions

list_diff = include(got, expected)

• got - structure to check
• expected - structure that is required in 'got' structure. Any difference will be in returned list.
• list_diff - list with differences, e.g. ['/attributes/temperature/<20.1 vs 22.2>']

Try to check if expected structure includes got structure. Any additional keys from got will be igroned. This situation can be expected if you don't want to check some parts of complex structure (e.g. in tests).

The structures should contain dictionaries, lists, objects, simple types or any comparable structures (__str__ and __eq__ implementation). Additionally expected can be or can contain compiled regular expression (re package) to check e.g. if you don't want to mock datetime objects.

Example of usage

So let's try test some endpoint:

import re
re_datetime = re.compile(r"^\d{4}-\d\d-\d\d \d\d:\d\d:\d\d$")

def test_add_new(self, client):
    with client:
        res = client.post(
            '/air_state',
            json = {
                'data': {
                    'type': 'air_state',
                    'attributes': {
                        'temperature': 20.1,
                        'humidity': 51.2,
                        'location': 'kitchen',
                        'device': 'dev1_esp',
                    },
                },
            },
            content_type = 'application/vnd.api+json',
        )
        assert 201 == res.status_code
        res_json = res.get_json()
        
        assert res_json['data']['type'] == 'air_state'
        assert res_json['data']['id'] == '1'
        assert res_json['data']['attributes']['temperature'] == '20.1'
        assert res_json['data']['attributes']['humidity'] == '51.2'
        assert res_json['data']['attributes']['location'] == 'kitchen'
        assert res_json['data']['attributes']['device'] == 'dev1_esp'
        assert re_datetime.search(res_json['data']['attributes']['created'])

You need to write a bunch of asserts for many items in result structure. Many lines of code. The bigger structure the more code.

Now you can write it other way using smart_data package:

from smart_data import include
from re import compile

def test_add_new(self, client):

    payload = {
        'type': 'air_state',
        'attributes': {
            'temperature': 20.1,
            'humidity': 51.2,
            'location': 'kitchen',
            'device': 'dev1_esp',
        },
    }

    with client:
        res = client.post(
            '/air_state',
            json = { 'data': payload },
            content_type = 'application/vnd.api+json',
        )
        assert 201 == res.status_code

        payload['attributes']['id'] = 1
        payload['attributes']['created'] = compile(r"^\d{4}-\d\d-\d\d \d\d:\d\d:\d\d$")
        res_json = res.get_json()
        assert include(got=res_json['data'], expected=payload) == []

This is a simple example with a really small amount of data to test. For more complex structures the benefit is higher.

The next benefit is readable output from broken assert during tests. For example:

    def test_foo():
        expected = {
            'foo': 1.1,
            'bar': [42, {'baz': 22}],
            'zoo': None,
            'zar': [[1, 3], [5, 8]],
        }
        got = {
            'foo': 1.1,
            'bar': [42, {'baz': 2}],
            'zoo': None,
            'zar': [[1, 3], [5, 8]],
        }
>       assert include(got, expected) == []
E       AssertionError: assert ['/bar/1/baz/<2 vs 22>'] == []
E         Left contains one more item: '/bar/1/baz/<2 vs 22>'
E         Full diff:
E         - []
E         + ['/bar/1/baz/<2 vs 22>']

tests/test_include.py:38: AssertionError

Enjoy!