A simple micro BDD framework for Python
Project description
GivenPy
GivenPy is a super simple yet highly extensible micro testing library to enable easier behavior driven development in Python.
It doesn't require any fancy dependencies and can work with any test execution framework like unittest or pytest.
It is basically a syntax sugar that allows you to write tests in a more readable way. It also helps you to encourage the re-use of setup code (given steps).
Examples
A quick preview of what you can do with GivenPy:
import unittest
from typing import Dict
from givenpy import given, when, then
from hamcrest import *
from starlette.testclient import TestClient
from tests.steps import prepare_api_server, create_test_client, prepare_injector
class TestExample(unittest.TestCase):
def test_health_works(self):
with given([
prepare_injector(), # custom code to prepare the test environment
prepare_api_server(),
create_test_client(),
]) as context:
client: TestClient = context.client
with when():
response: Dict = client.get('/api/v1/health').json()
with then():
assert_that(response, instance_of(dict))
assert_that(response['status'], is_(equal_to('OK')))
Installation
pip install givenpy PyHamcrest
Documentation
from givenpy import given, when, then
from hamcrest import *
def magic_function(x, external_number):
return x + 1 + external_number
def there_is_external_number():
context.external_number = 1
def test_magic_function():
with given([
there_is_external_number,
]) as context:
with when("I call the magic function"):
result = magic_function(1, context.external_number)
with then():
assert_that(result, is_(equal_to(3)))
But I recommend using more flexible higher functiosn that can become configurable:
from givenpy import given, when, then
from hamcrest import *
def magic_function(x, external_number):
return x + 1 + external_number
def there_is_external_number(number):
def step(context):
context.external_number = number
return step
def test_magic_function():
with given([
there_is_external_number(5),
]) as context:
with when("I call the magic function"):
result = magic_function(1, context.external_number)
with then("it should return 7"):
assert_that(result, is_(equal_to(7)))
You can also set up cleanup steps that will be executed after the test is finished:
from dataclasses import dataclass
from givenpy import given, when, then, lambda_with
from hamcrest import *
@dataclass
class Database:
connected: bool = False
def connect(self):
self.connected = True
def disconnect(self):
self.connected = False
def database_is_ready():
def step(context):
def open():
context.database = Database()
context.database.connect()
def close():
context.database.disconnect()
return lambda_with(open, close)
return step
def test_database_should_be_closed_when_we_exit_context():
with given([
database_is_ready(),
]) as context:
with then("it should connect to the database"):
assert_that(context.database.connected, is_(True))
with then("when context closes, it should disconnect from the database"):
assert_that(context.database.connected, is_(False))
Guidelines
- Always be explicit with your environment set up by adding all the necessary steps to the
given
block.- Make the setup steps as simple as possible. You can create master steps that will call other steps.
- Always prefer higher-order functions over simple functions for setup steps. This will allow you to create more flexible steps by allowing to parametrize them later.
- The test should always have given, when and then blocks.
- The test name should describe a functional behavior and should give an overview of what's happening and what are the
expectations.
- For example:
test_user_should_be_able_to_login
ortest_user_should_not_be_able_to_login_with_invalid_credentials
- And not:
test_login
ortest_invalid_credentials
- For example:
- The test should be readable and should not require any additional comments to understand what's happening.
- There should be only one
when
block per test. If you need multiple blocks, then you probably need multiple tests. - The
when
block should be as simple as possible. It should only contain the code that is being tested. Test the code from the end-user perspective. - Use PyHamcrest to write assertions. It's much more readable than the default unittest assertions.
- Do not be afraid to introduce custom matchers if needed.
- Extract complex & nested matchers to behavior-named functions that return the final matcher.
More examples
This example is from one of the bigger projects where we use givenpy to test our API endpoints.
import logging
import ulid
from hamcrest import *
from starlette.testclient import TestClient
from app.organizations.repository import TeamRepository
from app.organizations.team.core import Team
from givenpy import given, when, then
from tests.integration.organization.test_feedback_submission import person_is_present
from tests.integration.steps_auth import auth_is_ready
from tests.integration.steps_database import database_repo_is_ready, database_is_clean
from tests.integration.steps_issues import there_is_organization
from tests.steps import prepare_api_server, create_test_client, prepare_injector
logging.basicConfig(level=logging.DEBUG)
def test_team_creation_for_the_organization_should_work():
with given([
prepare_injector(),
database_repo_is_ready(),
database_is_clean(),
prepare_api_server(),
there_is_organization(),
create_test_client(),
auth_is_ready(),
]) as context:
client: TestClient = context.client
organization_id: ulid.ULID = context.organization_id
with when():
payload = {
"command_name": "CreateTeamCommand",
"entity_type": "team",
"payload": {
"name": "Test Team",
"organization_id": str(organization_id),
"members": [
]
}
}
response = client.post(
f'/api/v1/organizations/{organization_id}/teams',
headers=context.add_token(),
json=payload
)
with then():
assert_that(response.status_code, equal_to(200))
team_repo: TeamRepository = context.injector.get(TeamRepository)
team = team_repo.find_one(response.json()['id'])
assert_that(team.name, equal_to("Test Team"))
assert_that(team.organization_id, equal_to(organization_id))
def there_is_team(name="Test Team"):
def step(context):
team_repo: TeamRepository = context.injector.get(TeamRepository)
team = team_repo.save(
Team(
name=name,
organization_id=context.organization_id,
id=ulid.new(),
)
)
context.team_id = team.id
return step
def test_i_should_be_able_to_add_the_person_to_a_team():
with given([
prepare_injector(),
database_repo_is_ready(),
database_is_clean(),
prepare_api_server(),
there_is_organization(),
create_test_client(),
person_is_present(),
there_is_team("Test Team 1"),
auth_is_ready(),
]) as context:
client: TestClient = context.client
organization_id: ulid.ULID = context.organization_id
team_repo: TeamRepository = context.injector.get(TeamRepository)
with when():
payload = {
"command_name": "AddPersonToTeamCommand",
"entity_type": "team",
"payload": {
"team_id": str(context.team_id),
"person_id": str(context.person_id),
}
}
response = client.post(
f'/api/v1/organizations/{organization_id}/teams/{context.team_id}',
headers=context.add_token(),
json=payload
)
with then():
team = team_repo.find_one(context.team_id)
assert_that(response.status_code, equal_to(200))
assert_that(team.people, has_item(context.person_id))
def test_a_list_of_teams_for_the_current_organization_should_be_retrievable():
with given([
prepare_injector(),
database_repo_is_ready(),
database_is_clean(),
prepare_api_server(),
there_is_organization(),
create_test_client(),
person_is_present(),
there_is_team(name="Test Team 1"),
there_is_team("Test Team 1"),
auth_is_ready(),
]) as context:
client: TestClient = context.client
organization_id: ulid.ULID = context.organization_id
with when():
response = client.get(
f'/api/v1/organizations/{organization_id}/teams',
headers=context.add_token(),
)
with then():
assert_that(response.json(), has_length(2))
Contributing
Just create a PR or something. I'll review it and merge it if it's good.
Releasing a new version
git tag -a 1.0.2 -m "Tag 1.0.2"
git push
git push origin --tags
Project details
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
File details
Details for the file givenpy-1.0.3.tar.gz
.
File metadata
- Download URL: givenpy-1.0.3.tar.gz
- Upload date:
- Size: 5.0 kB
- Tags: Source
- Uploaded using Trusted Publishing? No
- Uploaded via: poetry/1.8.2 CPython/3.8.18 Linux/6.5.0-1015-azure
File hashes
Algorithm | Hash digest | |
---|---|---|
SHA256 | 5d87eeba69fb0f3a2ea4a35ab648f7ad9f6eb30eb45c9c214d48de48c85044c0 |
|
MD5 | 5225a1f6bb80f0bc99341f1f04fe69cf |
|
BLAKE2b-256 | 93757d9650217382a99986a1fc48cfb038dbabc1b617b21d9b6b6545a86ef2ba |
File details
Details for the file givenpy-1.0.3-py3-none-any.whl
.
File metadata
- Download URL: givenpy-1.0.3-py3-none-any.whl
- Upload date:
- Size: 5.3 kB
- Tags: Python 3
- Uploaded using Trusted Publishing? No
- Uploaded via: poetry/1.8.2 CPython/3.8.18 Linux/6.5.0-1015-azure
File hashes
Algorithm | Hash digest | |
---|---|---|
SHA256 | c8fb835ef0e1608018690eccb7ec59a9942679522b3715bc146b1a6191668d55 |
|
MD5 | a389fb9ffb559a88a756ef95ea8a07ff |
|
BLAKE2b-256 | 413ea7575becbf67e5bddd9e1eaaaba55426e639c0fe7a0844ec1465893d4d62 |