Let's test this code, with the obvious bug in it:

examples/python/unit_test/app.py

def add(x, y):
   return x*y

We can write a test-case like this:

examples/python/unit_test/test_app.py

import unittest
import app

class SomeTestCase(unittest.TestCase):
    def test_add(self):
        assert app.add(2, 2) == 4
        assert app.add(2, 3) == 5

Running with unittest.main

The "standard" or "default" way to run the test is by calling unittest.main as in this code:

examples/python/unit_test/with_main.py

import unittest
from test_app import SomeTestCase

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

It can be executed by running that file using python with_main.py

The result looks like this:

F
======================================================================
FAIL: test_add (test_app.SomeTestCase)
----------------------------------------------------------------------
Traceback (most recent call last):
  File "../examples/python/unit_test/test_app.py", line 7, in test_add
    assert app.add(2, 3) == 5
AssertionError

----------------------------------------------------------------------
Ran 1 test in 0.001s

FAILED (failures=1)

If we check the exit code it is

echo $?
1

Running TestSuite with unittest.TextTestRunner

examples/python/unit_test/with_text_test_runner.py

import unittest
from test_app import SomeTestCase


if __name__ == "__main__":
    suite = unittest.TestSuite()
    suite.addTest(SomeTestCase('test_add'))
    result = unittest.TextTestRunner(verbosity=2).run(suite)

    if result.wasSuccessful():
        exit(0)
    else:
        exit(1)

Just calling >unittest.TextTestRunner().run(suite) won't exit the script and thus won't set the exit code. This is not a good idea as external tools (e.g. CI systems) will rely on the exit code to determine if running the tests was successful or not.

So we capture the result of the test run in a variable and then call the wasSuccessful method on it to determine what should be the exit code.