How to make assertRaises() print the passed *args, **kwargs in Python 2.7?

Question

I'm using assertRaises in a loop like this:

for i in ['a', 'b', 'c']: 
    self.assertRaises(ValidationError, my_method, i)

And the problem is that whenever test fails, the output looks like this:

  File "/bla/test.py", line 4, in test_assertion
      my_method(i)
AssertionError: ValidationError not raised

How to make it print out the value of i when test fails? Like this:

  File "/bla/test.py", line 4, in test_assertion
      my_method('b')
AssertionError: ValidationError not raised

Answer 1

This is the ideal situation for the subTests context manager. However, this is only available in python 3. I think the best solution would be to create your own version of subTests. The benefits being it's easy to setup a basic mimicry of subTests, and if subTests is ever back ported to python 2 then it will be easy to switch.

import unittest
import sys
from contextlib import contextmanager

class TestCaseWithSubTests(unittest.TestCase):

    @contextmanager
    def subTest(self, **kwargs):
        try:
            yield None
        except:
            exc_class, exc, tb = sys.exc_info()
            kwargs_desc = ", ".join("{}={!r}".format(k, v) for k, v in kwargs.items())
            new_exc = exc_class("with {}: {}".format(kwargs_desc, exc))
            raise exc_class, new_exc, tb.tb_next

class MyTest(TestCaseWithSubTests):

    def test_with_assertion_error(self):
        for i in [0, 1]:
            with self.subTest(i=i), self.assertRaises(ZeroDivisionError):
                1 / i

    def test_with_value_error(self):
        def f(i):
            raise ValueError

        for i in [0, 1]:
            with self.subTest(i=i), self.assertRaises(ZeroDivisionError):
                f(i)


if __name__ == "__main__":
    unittest.main()

Which produces the following output:

FE
======================================================================
ERROR: test_with_value_error (__main__.MyTest)
----------------------------------------------------------------------
Traceback (most recent call last):
  File "C:\Users\dunes\py2sub_tests.py", line 30, in test_with_value_error
    f(i)
  File "C:\Python27\lib\contextlib.py", line 35, in __exit__
    self.gen.throw(type, value, traceback)
  File "C:\Users\dunes\py2sub_tests.py", line 30, in test_with_value_error
    f(i)
  File "C:\Users\dunes\py2sub_tests.py", line 26, in f
    raise ValueError
ValueError: with i=0: 

======================================================================
FAIL: test_with_assertion_error (__main__.MyTest)
----------------------------------------------------------------------
Traceback (most recent call last):
  File "C:\Users\dunes\py2sub_tests.py", line 22, in test_with_assertion_error
    1 / i
  File "C:\Python27\lib\contextlib.py", line 35, in __exit__
    self.gen.throw(type, value, traceback)
  File "C:\Users\dunes\py2sub_tests.py", line 22, in test_with_assertion_error
    1 / i
AssertionError: with i=1: ZeroDivisionError not raised

----------------------------------------------------------------------
Ran 2 tests in 0.006s

FAILED (failures=1, errors=1)

Answer 2

One solution is to use assertRaises() as a context manager.

with self.assertRaises(ValidationError):
    my_method(i)
    print("Failed when i was",i)

The print statement will only execute when the my_method() line passes successfully, so it will appear as the output when your test fails. Bit of a hack, but it works (unless my_method() throws some error that is not a ValidationError).