Next-gen Arrange-Act-Assert to structure test cases and force software engineers to express explicit intentions in
BDD style.
Table of content:
intentions is a Python next-gen Arrange-Act-Assert library created to help structuring test cases and force
software engineers to express explicit intentions in BDD style.
Arrange-Act-Assert is a pattern used to structure test cases. It provides a clear way to organize test code, making
it easier to read and understand:
Arrangeblock sets up the necessary objects, data, or state before the action to be tested.Actblock performs the actual operation or method call that should be tested.Assertblock checks the results of the action comparing the actual outcome with the expected outcome.
Behavior-Driven Development is an approach in testing that emphasizes the behavior of an application for business
needs. It focuses on defining test cases in plain, simple language with use cases and user stories. It typically uses
the Given-When-Then format to specify the system's expected behavior in various situations:
Givena context.Whenan action happens.Thenan expected outcome.
intentions aims to combine those two approaches to empower software engineers for more effective testing with when,
case, and expect context managers using which they can build behavior-driven arrange-act-assert based test cases.
In the same time you are not required to use it everywhere, even in the single test you are able to define which constructs to use and how many of them. For instance, you can skip it for unit tests and only use for integration tests.
class TestAccountService:
def test_transfer_money_with_insufficient_balance(self):
mock_send_in_app_notification = mock('notifications.send')
receiver_account = AccountFactory()
with when('Sender account has insufficient balance'):
sender_account = AccountFactory(balance=0)
with when('Pushing in-app notifications feature flag is enabled for sender account'):
enable_in_app_notifications(account=sender_account)
with case('Transfer money from one account to another'):
AccountService.transfer_money(from=sender_account, to=receiver_account, amount=100)
with expect('No transfers have been made'):
assert not sender_account.transfers
assert not receiver_account.transfers
with expect('Increase credit limit proposal is created'):
assert Proposal.get_last(account=sender_account, type=ProposalType.INCREASE_CREDIT_LIMIT)
with expect('Sender account receives insufficient balance in-app notification'):
mock_send_in_app_notification.assert_called_with(
account_id=sender_account.id,
type=InAppNotificationType.INSUFFICIENT_BALANCE,
expired_at=None,
)when, case and expect clearly convey the purpose of each block. With them, it is easy to tell a story of what is
being tested, making it easier for someone reading the test to understand the intention behind each part. Instead of
limited # Arrange, # Act and # Assert comments in your test case, you can use as much intentions as possible
emphasizing on every important detail of the test case.
With when, case and expect, test cases are broken down clearly, reducing mental load.
If the test fails, the when, case and expect make it easier to understand why the test was written in the first
place due to its descriptive nature. As the test case evolves, it's easier to maintain because the purpose of each step
is clear.
It aligns with Behavior-Driven Development principles, as it focuses on describing the behavior of the system under a
test case, making the tests more focused on outcomes and behavior rather than implementation details.
With when, case and expect, you introduce a common language for communication between developers, testers, and
business stakeholders and make collaboration easier. It also helps to minimize the learning curve for new joiners.
Install the library with the following command using pip3:
$ pip3 install intentionsIt emphasizes on setting up the necessary objects, data, or state to make the context of a test cace meaningful. Important to use this construct to focus exactly specific condition of the test case.
As you see on the example below, the context manager is used only on specifically a user from the UK that has
uploaded a document for a verification:
from intentions import when
class TestDocumentVerificationService:
def test_verify_document_when_uploaded_and_user_from_uk(self):
admin = UserFactory(is_admin=True)
verification = VerificationFactory()
...
with when('User is from the United Kingdom'):
user = UserFactory(country=Country.UK)
with when('User document is uploaded'):
user_document = Document(
user=user,
verification=verification,
status=DocumentStatus.UPLOADED,
)
...It emphasizes on performing the actual operation or method call that should be tested. Important to use this context manager over exact execution.
As you see on the example below, the context manager is used only for the document verification method named
verify_document. Besides having many other functions such as mocks, data preparation and side functions in the test
alongside the method:
from intentions import case
class TestDocumentVerificationService:
def test_verify_document_when_uploaded_and_user_from_uk(self):
...
mock_verification_api_response = prepare_verification_api_response()
mock_document_verification_api_request = mock(
path='api.document_verification.request',
data=mock_verification_api_response,
)
enable_async_requests()
create_user_kyc_profile()
with case('Verify a document'):
document_verification = DocumentVerificationService.verify_document(
document=document,
user=user,
admin=admin,
)
...It emphasizes on checking the results of the action comparing the actual outcome with the expected outcome, expected behavior or change in the system. Important to use this context manager to emphasize different groups of expectations and exact outcome or behavior.
As you see on the example below, the context manager is used to distinguish 3 different outcomes: the user's document was reviewed by external API, was additionally reviewed by admin as user is from the UK and the
document verification happened without errors:
from intentions import expect
class TestDocumentVerificationService:
def test_verify_document_when_uploaded_and_user_from_uk(self):
...
with expect('Document was verified by external API'):
mock_document_verification_api_request.assert_called()
assert document.is_verified_by_provider
with expect('Document was additionally verified by admin as user is from the UK'):
assert document.admin == admin
assert document.is_verified_by_admin
with expect('Document verification happened without errors'):
assert not document_verification.errors