def test_with_block_closes_runner_after_error():
with expect.raises(KeyError):
with lola.Runner() as runner:
raise KeyError()
with expect.raises(ValueError):
runner.close()
def test_call_with_closed_runner_throws_error():
runner = lola.Runner()
runner.close()
with expect.raises(ValueError):
runner.call(['test_scripts/success.py'],
stdout=DEVNULL,
stderr=DEVNULL)
def test_cwd():
with temp_directory() as tmpdir:
with lola.Runner() as runner:
out = runner.check_output([os.path.abspath('test_scripts/pwd.py')],
cwd=tmpdir)
# use samefile instead of string comparison since on OSX we get "/var" as
# tmpdir, but "/private/var" from the pwd script
assert os.path.samefile(out, tmpdir)
def test_env_inherit_from_parent():
# when "env" is not set explicitly this should be passed through
os.environ['LOLA_ENV_TEST_PARENT'] = 'set in parent'
with lola.Runner() as runner:
out = runner.check_output(
['test_scripts/env.py', 'LOLA_ENV_TEST_PARENT'])
expect(out) == 'set in parent'
def test_env():
env = os.environ.copy()
env['LOLA_ENV_TEST'] = 'test value'
with lola.Runner() as runner:
out = runner.check_output(['test_scripts/env.py', 'LOLA_ENV_TEST'],
env=env)
expect(out) == 'test value'
def test_check_output_error():
try:
with lola.Runner() as runner:
runner.check_output(['test_scripts/fail.py'],
stderr=subprocess.STDOUT)
except subprocess.CalledProcessError as e:
expect(e.returncode) == 1
expect(e.output) == 'failing'
else:
assert False, 'Should have raise a CalledProcessError'
def test_stderr_to_stdout():
with temp_directory() as tmpdir:
outfile = os.path.join(tmpdir, 'out.txt')
with lola.Runner() as runner:
runner.check_call(['test_scripts/output.py'],
stdout=open(outfile, 'w'),
stderr=subprocess.STDOUT)
expect(open(outfile).read()) == 'outputting\nerroring\n'
def test_check_call_error():
try:
with lola.Runner() as runner:
runner.check_call(['test_scripts/fail.py'],
stdout=DEVNULL,
stderr=DEVNULL)
except subprocess.CalledProcessError as e:
expect(e.returncode) == 1
else:
assert False, 'Should have raise a CalledProcessError'
def test_env_overrides_parent():
env = os.environ.copy()
# this key set in the parent process should not be inherited when
# env is passed
os.environ['LOLA_ENV_TEST_PARENT'] = 'set in parent'
with lola.Runner() as runner:
out = runner.check_output(
['test_scripts/env.py', 'LOLA_ENV_TEST_PARENT'], env=env)
expect(out) == 'not set'
def test_output_to_file():
with temp_directory() as tmpdir:
outfile = os.path.join(tmpdir, 'out.txt')
errfile = os.path.join(tmpdir, 'err.txt')
with lola.Runner() as runner:
runner.check_call(['test_scripts/output.py'],
stdout=open(outfile, 'w'),
stderr=open(errfile, 'w'))
expect(open(outfile).read()) == 'outputting\n'
expect(open(errfile).read()) == 'erroring\n'
def test_check_output():
with lola.Runner() as runner:
out = runner.check_output(['test_scripts/output.py'],
stderr=subprocess.STDOUT)
expect(out) == 'outputting\nerroring\n'
def test_check_call_success():
with lola.Runner() as runner:
retcode = runner.check_call(['test_scripts/success.py'],
stdout=DEVNULL,
stderr=DEVNULL)
expect(retcode) == 0
def test_exception():
with lola.Runner() as runner:
retcode = runner.call(['test_scripts/exception.py'],
stdout=DEVNULL,
stderr=DEVNULL)
expect(retcode) == 255
def test_failure_retcode():
with lola.Runner() as runner:
retcode = runner.call(['test_scripts/fail.py'],
stdout=DEVNULL,
stderr=DEVNULL)
expect(retcode) == 1
def test_with_block_closes_runner():
with lola.Runner() as runner:
pass
with expect.raises(ValueError):
runner.close()
def test_reclose_throws_error():
runner = lola.Runner()
runner.close()
with expect.raises(ValueError):
runner.close()
