kcarnold / zylabs-libs

Utility code for ZyBooks autograders in Python

zylabs-libs

Utility code for ZyBooks autograders in Python

test_util.py

Add test_util.py as an "additional file" in the test. Then just add a decorator to your test_passed function for instant upgrades:

• Turns AssertionErrors into test feedback
• Catches syntax and other errors
• Allows providing arbitrary input (and fails if input() is called but no input was provided)
• Captures standard output.
• Has the test fail on any errors, pass otherwise.
• Interprets None returns as passing.

Example unit test:

from test_util import test

@test()
def test_passed(test_feedback):
    from fraction import Fraction
    f1 = Fraction(1, 2)
    f2 = Fraction(2, 3)
    f3 = f1 * f2
    assert isinstance(f3, Fraction), "__mul__ should return a Fraction."
    assert str(f3) == '1/3', "1/2 * 2/3 should equal 1/3."
    assert f1.numerator == 1 and f1.denominator == 2, "__mul__ should not change 'self'."
    assert f2.numerator == 2 and f2.denominator == 3, "__mul__ should not change the other fraction."

doctests

Students sometimes have trouble with the unit tests because they don't know how to reproduce a failure. The Python standard library has a solution for this: doctests. I've been trying this out with ZyLabs; see doctest_example.py for an example of how to write a unit test that's actually a doctest. Students need some coaching to be able to interpret the doctest error report, especially since I haven't yet cleaned up the stack trace in the doctest report like I do in @test().

It'll return a fractional score based on how many of the doctests passed.

Note that it doesn't distinguish between returning a value and printing it. So you may need to demo the effect of that, like:

>>> not is_prime(4)
True

I don't think it'll run any doctests that students write, but that would be a nifty extension (to be able to test their tests, for example).