Skip to main content

Allows you to run a test with multiple data sets

Project description

http://opensource.box.com/badges/active.svg https://travis-ci.org/box/genty.png?branch=master https://img.shields.io/pypi/v/genty.svg https://img.shields.io/pypi/dm/genty.svg

About

Genty, pronounced “gen-tee”, stands for “generate tests”. It promotes generative testing, where a single test can execute over a variety of input. Genty makes this a breeze.

For example, consider a file sample.py containing both the code under test and the tests:

from genty import genty, genty_repeat, genty_dataset
from unittest import TestCase

# Here's the class under test
class MyClass(object):
    def add_one(self, x):
        return x + 1

# Here's the test code
@genty
class MyClassTests(TestCase):
    @genty_dataset(
        (0, 1),
        (100000, 100001),
    )
    def test_add_one(self, value, expected_result):
        actual_result = MyClass().add_one(value)
        self.assertEqual(expected_result, actual_result)

Running the MyClassTests using the default unittest runner

$ python -m unittest -v sample
test_add_one(0, 1) (sample.MyClassTests) ... ok
test_add_one(100000, 100001) (sample.MyClassTests) ... ok

----------------------------------------------------------------------
Ran 2 tests in 0.000s

OK

Instead of having to write multiple independent tests for various test cases, code can be refactored and parametrized using genty!

It produces readable tests. It produces maintainable tests. It produces expressive tests.

Another option is running the same test multiple times. This is useful in stress tests or when exercising code looking for race conditions. This particular test

@genty_repeat(3)
def test_adding_one_to_zero(self):
    self.assertEqual(1, MyClass().add_one(0))

would be run 3 times, producing output like

$ python -m unittest -v sample
test_adding_one() iteration_1 (sample.MyClassTests) ... ok
test_adding_one() iteration_2 (sample.MyClassTests) ... ok
test_adding_one() iteration_3 (sample.MyClassTests) ... ok

----------------------------------------------------------------------
Ran 3 tests in 0.001s

OK

The 2 techniques can be combined:

@genty_repeat(2)
@genty_dataset(
    (0, 1),
    (100000, 100001),
)
def test_add_one(self, value, expected_result):
    actual_result = MyClass().add_one(value)
    self.assertEqual(expected_result, actual_result)

There are more options to explore including naming your datasets and genty_args.

@genty_dataset(
    default_case=(0, 1),
    limit_case=(999, 1000),
    error_case=genty_args(-1, -1, is_something=False),
)
def test_complex(self, value1, value2, optional_value=None, is_something=True):
    ...

would run 3 tests, producing output like

$ python -m unittest -v sample
test_complex(default_case) (sample.MyClassTests) ... ok
test_complex(limit_case) (sample.MyClassTests) ... ok
test_complex(error_case) (sample.MyClassTests) ... ok

----------------------------------------------------------------------
Ran 3 tests in 0.003s

OK

The @genty_datasets can be chained together. This is useful, for example, if there are semantically different datasets so keeping them separate would help expressiveness.

@genty_dataset(10, 100)
@genty_dataset('first', 'second')
def test_composing(self, parameter_value):
        ...

would run 4 tests, producing output like

$ python -m unittest -v sample
test_composing(10) (sample.MyClassTests) ... ok
test_composing(100) (sample.MyClassTests) ... ok
test_composing(u'first') (sample.MyClassTests) ... ok
test_composing(u'second') (sample.MyClassTests) ... ok

----------------------------------------------------------------------
Ran 4 tests in 0.000s

OK

Sometimes the parameters to a test can’t be determined at module load time. For example, some test might be based on results from some http request. And first the test needs to authenticate, etc. This is supported using the @genty_dataprovider decorator like so:

def setUp(self):
    super(MyClassTests, self).setUp()

    # http authentication happens
    # And imagine that _some_function is actually some http request
    self._some_function = lambda x, y: ((x + y), (x - y), (x * y))

@genty_dataset((1000, 100), (100, 1))
def calculate(self, x_val, y_val):
    # when this is called... we've been authenticated
    return self._some_function(x_val, y_val)

@genty_dataprovider(calculate)
def test_heavy(self, data1, data2, data3):
    ...

would run 4 tests, producing output like

$ python -m unittest -v sample
test_heavy_calculate(100, 1) (sample.MyClassTests) ... ok
test_heavy_calculate(1000, 100) (sample.MyClassTests) ... ok

----------------------------------------------------------------------
Ran 2 tests in 0.000s

OK

Notice here how the name of the helper (calculate) is added to the names of the 2 executed test cases.

Like @genty_dataset, @genty_dataprovider can be chained together.

Enjoy!

Deferred Parameterization

Parametrized tests where the final parameters are not determined until test execution time. It looks like so:

@genty_dataset((1000, 100), (100, 1))
def calculate(self, x_val, y_val):
    # when this is called... we've been authenticated, perhaps in
    # some Test.setUp() method.

    # Let's imagine that _some_function requires that authentication.
    # And it returns a 3-tuple, matching that signature of
    # of the test(s) decorated with this 'calculate' method.
    return self._some_function(x_val, y_val)

@genty_dataprovider(calculate)
def test_heavy(self, data1, data2, data3):
    ...

The calculate() method is called 2 times based on the @genty_dataset decorator, and each of it’s return values define the final parameters that will be given to the method test_heavy(...).

Installation

To install, simply:

pip install genty

Contributing

See CONTRIBUTING.rst.

Setup

Create a virtual environment and install packages -

mkvirtualenv genty
pip install -r requirements-dev.txt

Testing

Run all tests using -

tox

The tox tests include code style checks via pep8 and pylint.

The tox tests are configured to run on Python 2.6, 2.7, 3.3, 3.4, 3.5, and PyPy 2.6.

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

genty-1.3.2.tar.gz (20.8 kB view hashes)

Uploaded Source

Built Distribution

genty-1.3.2-py2.py3-none-any.whl (19.3 kB view hashes)

Uploaded Python 2 Python 3

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page