The py.test testing tool makes it easy to write small tests, yet scales to support complex functional testing. It provides

auto-discovery of test modules and functions,
detailed info on failing assert statements (no need to remember self.assert* names)
modular fixtures for managing small or parametrized long-lived test resources.
multi-paradigm support: you can use py.test to run test suites based on unittest (or trial), nose
single-source compatibility to Python2.4 all the way up to Python3.3, PyPy-1.9 and Jython-2.5.1.
many external plugins.

A simple example for a test:

# content of test_module.py
def test_function():
    i = 4
    assert i == 3

which can be run with py.test test_module.py. See getting-started for more examples.

For much more info, including PDF docs, see

http://pytest.org

and report bugs at:

http://bitbucket.org/hpk42/pytest/issues/

and checkout repos at:

http://github.com/hpk42/pytest/ (mirror) http://bitbucket.org/hpk42/pytest/