django-testdata 1.0

Installation

pip install django-testdata

Motivation

Django 1.8 introduced TestCase.setUpTestData to allow costly generation of model fixtures to be executed only once per test class in order to speed up testcase instances execution.

One gotcha of setUpTestData though is that test instances all share the same model instances and have to be careful not to alter them to prevent breaking test isolation. Per Django’s documentation:

Be careful not to modify any objects created in setUpTestData() in your
test methods. Modifications to in-memory objects from setup work done at
the class level will persist between test methods. If you do need to modify
them, you could reload them in the setUp() method with refresh_from_db(),
for example.

Reloading objects in setUp() certainly works but it kind of defeats the purpose of avoiding database hits to speed up tests execution in the first place. It makes little sense to fetch model instances from the database given all

This package offers a different alternative to work around this quirk of setUpTestData. Instead of reloading objects from the database the model instances assigned as class attributes during setUpTestData are lazily deep copied on testcase instance accesses from their original definition. All of deep copying is done by sharing a memo which makes sure in-memory relationships between objects is preserved.

Usage

The test data can be either wrapped manually by using testdata.

from django.test import TestCase
from testdata import testdata

from .models import Author, Book

class BookTests(TestCase):
    @classmethod
    def setUpTestData(cls):
        cls.author = testdata(Author.objects.create(
            name='Milan Kundera',
        ))
        cls.book = testdata(cls.author.books.create(
            title='Nesnesitelná lehkost bytí',
        ))

Or automatically by using the wrap_testdata decorator.

from django.test import TestCase
from testdata import testdata

from .models import Author, Book

class BookTests(TestCase):
    @classmethod
    @wrap_testdata
    def setUpTestData(cls):
        cls.author = Author.objects.create(
            name='Milan Kundera',
        )
        cls.book = cls.author.books.create(
            title='Nesnesitelná lehkost bytí',
        )

Under the hood wrap_testdata simply wraps all attributes added to cls during the execution of setUpTestData() into testdata(attr, name=name) which has also the nice side effect of speeding subsequent accesses.

Once test data is wrapped the testcase instances methods can alter objects retrieved from self without worrying about cross-tests isolation.

from django.test import TestCase
from testdata import testdata

from .models import Author, Book

class BookTests(TestCase):
    @classmethod
    @wrap_testdata
    def setUpTestData(cls):
        cls.author = Author.objects.create(
            name='Milan Kundera',
        )
        cls.book = cls.author.books.create(
            title='Nesnesitelná lehkost bytí',
        )

    def test_book_name_english(self):
        self.assertEqual(self.book.title, 'Nesnesitelná lehkost bytí')
        self.book.title = 'The Unbearable Lightness of Being'
        self.book.save()

    def test_book_name_french(self):
        self.assertEqual(self.book.title, 'Nesnesitelná lehkost bytí')
        self.book.title = "L'Insoutenable Légèreté de l'être"
        self.book.save()