BDD library for the pytest runner

pytest-bdd-ng combine descriptive clarity of Gherkin language with power and fullness of pytest infrastructure. It enables unifying unit and functional tests, reduces the burden of continuous integration server configuration and allows the reuse of test setups.

Pytest fixtures written for unit tests can be reused for setup and actions mentioned in feature steps with dependency injection. This allows a true BDD just-enough specification of the requirements without obligatory maintaining any context object containing the side effects of Gherkin imperative declarations.

Why NG ?

The current pytest plugin for cucumber is pytest-bdd , a popular project with 1.2k stars and used in 3k public repos and maintained by the pytest community. The upstream open-cucumber project does not have an official python release, so the current cucumber specs include features not available in pytest-bdd . This project is an effort to bridge the gap and also make it easier for pytest users to access new cucumber features.

Feature	original	NG	Description
Official parser support	-	+	All features of Feature files are supported from the "box" (localisation, Rules, Examples, Data tables, etc.)
Steps definitions via Cucumber expressions	-	+	Easy migration between implementations
Reporting using Messages	-	+	Possible to use all collection of Cucumber community tools for reporting
Pickles internal protocol	-	+	Allows to implement parsers based on other file types/principles
Heuristic step matching	-/+	+	Steps ease of use / amount of needed boilerplate code
Step execution context. Step types and variants of definition	-/+	+	Dispatching steps by kind. Steps injecting multiple fixtures. Default injecting behaviors. Steps could be used on import automatically. It's possible to define default values for step variables.
Automatic collection of Feature files	-	+	No boilerplate code / No mix between steps definition and feature files
Load of Feature files by HTTP	-	+	Allows to integrate the library into external Feature storages
Stability and bugfixes	+	+/-
Supported python/pytest versions	+/-	+	NG supports wider and elder pytest&python version. Tested also for PyPy
Active community	+	-/+

Install pytest-bdd-ng

pip install pytest-bdd-ng

Packages for parsing Markdown defined features

npm install @cucumber/gherkin

Packages for reporting

npm install @cucumber/html-formatter

TLDR;

Feature: Simplest example
  Scenario:
    # https://cucumber.io/docs/gherkin/reference/
    Given File "Passing.feature" with content:
      """gherkin
      Feature: Passing feature
        Scenario: Passing scenario
          * Passing step
      """
    # https://docs.pytest.org/en/stable/reference/fixtures.html#conftest-py-sharing-fixtures-across-multiple-files
    And File "conftest.py" with content:
      """python
      from pytest_bdd import step

      @step('Passing step')
      def _():
        ...
      """
    # https://docs.pytest.org/en/stable/how-to/usage.html
    When run pytest
    Then pytest outcome must contain tests with statuses:
      |passed|
      |     1|

Project layout

pytest-bdd-ng automatically collects *.feature files from pytest tests directory. Important to remember, that feature files are used by other team members as live documentation, so it's not a very good idea to mix documentation and test code.

The more features and scenarios you have, the more important becomes the question about their organization. So the recommended way is to organize your feature files in the folders by semantic groups:

features
├──frontend
│  └──auth
│     └──login.feature
└──backend
   └──auth
      └──login.feature

And tests for these features would be organized in the following manner:

tests
└──conftest.py
└──functional
│     └──__init__.py
│     └──conftest.py
│     │     └── "User step library used by descendant tests"
│     │
│     │         from steps.auth.given import *
│     │         from steps.auth.when import *
│     │         from steps.auth.then import *
│     │
│     │         from steps.order.given import *
│     │         from steps.order.when import *
│     │         from steps.order.then import *
│     │
│     │         from steps.browser.given import *
│     │         from steps.browser.when import *
│     │         from steps.browser.then import *
│     │
│     └──frontend_auth.feature -> ../../features/frontend/auth.feature
│     └──backend_auth.feature -> ../../features/backend/auth.feature
...

The step definitions would then be organized like this:

steps
└──auth
│     └── given.py
│     │      └── """User auth step definitions"""
│     │          from pytest import fixture
│     │          from pytest_bdd import given, when, then, step
│     │
│     │          @fixture
│     │          def credentials():
│     │             return 'test_login', 'test_very_secure_pass'
│     │
│     │          @given('User login into application')
│     │          def user_login(credentials):
│     │             ...
│     └── when.py
│     └── then.py
└──order
│     └── given.py
│     └── when.py
│     └── then.py
└──browser
│     └── ...
...

To make links between feature files at features directory and test directory there are few options (for more information please examine the project's tests):

1. Symlinks
2. .desktop files
3. .webloc files
4. .url files

Note

Link files also could be used to load features by http://

How to Contribute

The project is now open to contributions. Please open an issue for more details.