pytest-beehave 1.0.0

Generates pytest test stubs from Gherkin .feature files, checks consistency, and displays BDD steps — automatically, every time pytest runs.

pytest-beehave is a pytest plugin powered by beehave. It reads your Gherkin .feature files and generates Hypothesis-compatible test stubs with the right names, decorators, and parameters — then verifies that your test code stays consistent with your spec. New scenarios get @pytest.mark.skip; drift becomes a real test failure.

pip install pytest-beehave

No conftest.py changes required. The plugin registers itself via pytest's entry-point system.

Optional: install pytest-beehave[html] for a "Scenario" column in pytest-html reports.

pip install "pytest-beehave[html]"

Quick start

1. Write a feature file

# docs/features/checkout/shopping_cart.feature
Feature: Shopping cart

  Background:
    Given an empty cart

  Rule: Tax calculation

    Scenario: VAT is applied at the correct rate
      Given a cart with items totalling $100
      When the buyer is in the UK
      Then the order total is $120

2. Run pytest

pytest --collect-only

3. A test stub is created

tests/features/shopping_cart/
├── tax_calculation_test.py

# tests/features/shopping_cart/tax_calculation_test.py
import pytest

@pytest.mark.skip(reason="not implemented")
def test_VAT_is_applied_at_the_correct_rate():
    ...

4. See BDD steps

pytest -v

tests/features/shopping_cart/tax_calculation_test.py::test_VAT_is_applied_at_the_correct_rate SKIPPED
  Given an empty cart
  Given a cart with items totalling $100
  When the buyer is in the UK
  Then the order total is $120

5. Implement and ship

Remove @pytest.mark.skip, write the test body, run pytest again. The steps display stays in sync with your feature file.

How it works

The plugin hooks into pytest_configure — every stub exists on disk before collection begins.

pytest invoked
  └─ pytest_configure fires
       ├─ load_config()                  → read [tool.beehave] from pyproject.toml
       ├─ parse_feature()                → parse .feature files into ScenarioInfo
       ├─ generate_stubs()               → write Hypothesis test stubs to disk
       ├─ _add_skip_markers()            → mark unimplemented stubs with @pytest.mark.skip
       ├─ check_all()                    → detect drift between features and tests
       └─ register display plugins       → StepsReporter (-v) and/or HtmlStepsPlugin
  └─ pytest_collection_modifyitems       → inject failing tests for ERROR violations
  └─ Collection begins — every stub is already present

File layout

docs/features/              ← configured via features_dir
  **/*.feature              ← any subfolder structure is supported

tests/features/             ← configured via tests_dir
  <feature_slug>/           ← one directory per feature (derived from Feature title)
    <rule_slug>_test.py     ← one file per Rule: block (or default_test.py)

Each test function name follows test_<scenario_title_with_underscores>. The mapping is exact string equality — no @id tags, no step definitions, no glue code.

What check_all enforces

After stub generation, the plugin runs check_all() to detect drift between feature files and test code. ERROR violations produce real test failures via synthetic test items:

$ pytest
[beehave] ERROR: tests/features/demo/default_test.py:5: unmapped-test: 'test_orphan' has no matching scenario
========================= 1 failed, 3 passed, 2 skipped =========================

Stub functions (body is ...) are excluded from placeholder and literal checks.

How it maps

• Scenario title → function name: VAT Is Applied At The Correct Rate → test_VAT_is_applied_at_the_correct_rate. Lowercased. Globally unique.
• Rule → test file: Top-level scenarios go to default_test.py. Scenarios inside a Rule go to <rule>_test.py.
• Feature title → directory: Shopping Cart → tests/features/shopping_cart/.
• Scenario Outline → decorators: <placeholder> columns become @given() parameters with inferred Hypothesis strategies. Example rows become @example() decorators.

TDD workflow

1. pytest --collect-only → stubs generated with @pytest.mark.skip
2. Remove @pytest.mark.skip, write the test body → test runs and fails (red)
3. Fix the implementation → test passes (green)
4. Add new scenarios to .feature files → only new stubs get the skip marker

Configuration

All configuration lives under [tool.beehave] in pyproject.toml:

[tool.beehave]
features_dir = "docs/features"      # default: docs/features
tests_dir = "tests/features"        # default: tests/features
default_strategy = "text"           # default: text (Hypothesis strategy for placeholders)
background_check_numeric = true     # default: true
background_check_string = true      # default: true

If features_dir does not exist, the plugin exits silently.

Requirements

Contributing

git clone https://github.com/nullhack/pytest-beehave
cd pytest-beehave
uv sync --all-extras
uv run task test && uv run task lint && uv run task static-check

Bug reports and pull requests welcome on GitHub.

License

MIT — see LICENSE.