launch_testing

This tool is a framework for launch integration testing. For example:

• The exit codes of all processes are available to the tests.
• Tests can check that all processes shut down normally, or with specific exit codes.
• Tests can fail when a process dies unexpectedly.
• The stdout and stderr of all processes are available to the tests.
• The command-line used to launch the processes are available to the tests.
• Some tests run concurrently with the launch and can interact with the running processes.

Quick start example

Start with the launch_testing example good_proc_launch_test.py.

Run the example by doing:

launch_test test/launch_testing/examples/good_proc_launch_test.py

launch_test will launch the nodes found in the generate_test_description function, run the tests from the TestGoodProcess class, shut down the launched nodes, and then run the tests from the TestProcessOutput class.

The Launch Description

def generate_test_description():

    return launch.LaunchDescription([
        launch.actions.ExecuteProcess(
            cmd=[path_to_process],
        ),

        # Start tests right away - no need to wait for anything in this example.
        # In a more complicated launch description, we might want this action happen
        # once some process starts or once some other event happens
        launch_testing.actions.ReadyToTest()
    ])

The generate_test_description function should return a launch.LaunchDescription object that launches the system to be tested.

The launch description needs to include a ReadyToTest action to signal to the test framework that it’s safe to start the active tests.

In the above example, there is no need to delay the start of the tests so the ReadyToTest action is a peer to the process under test and will signal to the framework that it’s safe to start around the same time the ExecuteProcess action is run.

Active Tests

Any classes that inherit from unittest.TestCase and not decorated with the post_shutdown_test descriptor will be run concurrently with the process under test. These tests are expected to interact with the running processes in some way.

Post-Shutdown Tests

Any classes that inherit from unittest.TestCase that are decorated with the post_shutdown_test descriptor will be run after the launched processes have been shut down. These tests have access to the exit codes and the stdout of all of the launched processes, as well as any data created as a side-effect of running the processes.

Exit Codes and Standard Out

The launch_testing framework automatically adds some member fields to each test case so that the tests can access process output and exit codes.

• self.proc_info - a ProcInfoHandler object
• self.proc_output - an IoHandler object

These objects provide dictionary like access to information about the running processes. They also contain methods that the active tests can use to wait for a process to exit or to wait for specific output.

Assertions

The launch_testing framework automatically records all stdout from the launched processes as well as the exit codes from any processes that are launched. This information is made available to the tests via the proc_info and proc_output object. These objects can be used by one of several assert methods to check the output or exit codes of the process:

launch_testing.asserts.assertInStdout(proc_output, msg, process, cmd_args=None, *, strict_proc_matching=True)

Asserts that a message is found in the stdout of a particular process.

• msg:

  The text to look for in the process standard out

• process:

  Either the process name as a string, or a launch.actions.ExecuteProcess object that was used to start the process. Pass None or an empty string to search all processes.

• cmd_args:

  When looking up processes by process by name, cmd_args can be used to disambiguate multiple processes with the same name.

• strict_proc_matching:

  When looking up a process by name, strict_proc_matching=True will make it an error to match multiple processes. This prevents an assert from accidentally passing if the output came from a different process than the one the user was expecting.

launch_testing.asserts.assertExitCodes(proc_info, allowable_exit_codes=[EXIT_OK], process, cmd_args=None, *, strict_proc_matching=True)

Asserts that the specified processes exited with a particular exit code.

• allowable_exit_codes:

  A list of allowable exit codes. By default EXIT_OK (0) plus EXIT_FORCED (1) on Windows. Other exit codes provided are EXIT_SIGINT (130), EXIT_SIGQUIT (131), EXIT_SIGKILL (137) and EXIT_SIGSEGV (139).

• The process, cmd_args, and strict_proc_matching arguments behave the same way as in assertInStdout. By default, assert on the exit codes of all processes.

launch_testing.asserts.assertSequentialStdout(proc_output, process, cmd_args=None)

Asserts that standard out was seen in a particular order.

• process and cmd_args:

  These arguments are the same as in assertInStdout and assertExitCodes, however it is not possible to match multiple processes because there is no way to determine the order of stdout that came from multiple processes.

Returns a context manager that will check that a series of assertions happen in order.

As an example, consider:

with assertSequentialStdout(self.proc_output, "proc_name") as cm:
    cm.assertInStdout("Loop 1")
    cm.assertInStdout("Loop 2")
    cm.assertInStdout("Loop 3")

Waiting for Output or Exit Codes

The ActiveTests can also call methods that wait for particular output or a particular process to exit or time out. These asserts are methods on the proc_output and proc_info objects.

proc_output.assertWaitFor(msg, process=None, cmd_args=None, *, strict_proc_matching=True, timeout=10)

• msg, process, cmd_args, and strict_proc_matching:

  These arguments work the same as in other assert methods. By default, this method waits on output from any process.

• timeout:

  The amount of time to wait before raising an AssertionError.

proc_info.assertWaitForShutdown(process, cmd_args=None, *, timeout=10)

• process and cmd_args:

  These arguments work the same as in other assertions, but it is not possible to wait on multiple processes to shut down.

• timeout:

  The amount of time to wait before raising an AssertionError

Arguments

launch_test uses launch arguments for tests too.

Arguments are declared in the launch description and can be accessed by the test via a test_args dictionary that’s injected into the tests similar to proc_info and proc_output.

launch_test --show-args test/launch_testing/examples/args_launch_test.py
launch_test test/launch_testing/examples/args_launch_test.py dut_arg:=value

See the launch_testing example with arguments for further reference.

Using CMake

To run launch tests from a CMakeLists.txt file, you’ll need to declare a dependency on launch_testing_ament_cmake in your package.xml.

Then, in the CMakeLists.txt file, add:

find_package(launch_testing_ament_cmake)
add_launch_test(test/name_of_test.test.py)

Arguments can be passed to the tests via the CMake function, too:

add_launch_test(
  test/test_with_args.test.py
  ARGS "arg1:=foo"
)

Examples

hello_world_launch_test.py

Usage:

launch_test test/launch_testing/examples/hello_world_launch_test.py

This test is a simple example on how to use the launch_testing.

It launches a process and asserts that it prints “hello_world” to stdout using proc_output.assertWaitFor(). Finally, it checks if the process exits normally (zero exit code).

The @launch_testing.markers.keep_alive decorator ensures that the launch process stays alive long enough for the tests to run.

good_proc_launch_test.py

Usage:

launch_test test/launch_testing/examples/good_proc_launch_test.py

This test checks a process called good_proc.py (source found in the example_processes folder). good_proc.py is a simple python process that prints “Loop 1, Loop2, etc. every second until it’s terminated with ctrl+c. The test will launch the process, wait for a few loops to complete by monitoring stdout, then terminate the process and run some post-shutdown checks.

The pre-shutdown tests check that “Loop 1, Loop 2, Loop 3, Loop 4” are all printed to stdout. Once this test finishes, the process under test is shut down

After shutdown, we run a similar test that checks more output, and also checks the order of the output. test_out_of_order demonstrates that the assertSequentialStdout context manager is able to detect out of order stdout.

terminating_proc_launch_test.py

Usage:

launch_test test/launch_testing/examples/terminating_proc_launch_test.py

This test checks proper functionality of the terminating_proc example (source found in the example_processes folder).

args_launch_test.py

Usage to view the arguments:

launch_test test/launch_testing/examples/args_launch_test.py --show-args

Usage to run the test:

launch_test test/launch_testing/examples/args_launch_test.py dut_arg:=hey

This example shows how to pass arguments into a launch test. The arguments are made available in the launch description via a launch.substitutions.LaunchConfiguration. The arguments are made available to the test cases via a self.test_args dictionary

This example will fail if no arguments are passed.

context_launch_test.py

Usage:

launch_test test/launch_testing/examples/context_launch_test.py

This example shows how the generate_test_description function can return a tuple where the second item is a dictionary of objects that will be injected into the individual test cases. Tests that wish to use elements of the test context can add arguments with names matching the keys of the dictionary.

Changelog for package launch_testing

3.6.1 (2024-07-29)

3.6.0 (2024-06-25)

3.5.1 (2024-06-17)

• Add mechanism to disable workaround for dependency groups (#775)
• Contributors: Scott K Logan

3.5.0 (2024-04-26)

3.4.2 (2024-04-16)

• Fix a warning in modern unittest. (#773) Newer versions of unittest no longer store an errors list; instead, they store a result, which then stores an error list. Update the code here to be able to deal with either version.
• Contributors: Chris Lalancette

3.4.1 (2024-03-28)

• Add consider_namespace_packages=False (#766)
• Contributors: Tony Najjar

3.4.0 (2024-02-07)

3.3.0 (2024-01-24)

3.2.1 (2023-12-26)

3.2.0 (2023-10-04)

• to open expected outpout file with an encoding parameter (#717)
• Contributors: Chen Lihui

3.1.0 (2023-09-08)

3.0.1 (2023-09-07)

3.0.0 (2023-08-21)

2.2.1 (2023-07-11)

2.2.0 (2023-06-07)

2.1.0 (2023-04-27)

2.0.1 (2023-04-12)

2.0.0 (2023-04-11)

• Improve type checking (#679)
• Contributors: Hervé Audren

1.4.1 (2023-02-24)

• Fixed typos (#692)
• Contributors: Alejandro Hernández Cordero

1.4.0 (2023-02-14)

• Allow ReadyToTest() usage in event handler (#665)
• Inherit markers from generate_test_description (#670)
• [rolling] Update maintainers - 2022-11-07 (#671)
• Contributors: Audrow Nash, Nikolai Morin, Scott K Logan

1.3.0 (2022-11-02)

1.2.0 (2022-09-13)

• Fix Typo (#641)
• ReadyToTest action timeout using decorator (#625)
• Switch to using a comprehension for process_names. (#614)
• Contributors: Chris Lalancette, Deepanshu Bansal, Kenji Brameld

1.1.0 (2022-04-29)

1.0.1 (2022-04-13)

1.0.0 (2022-04-12)

• Removed the deprecated [ready_fn]{.title-ref} feature (#589)
• Contributors: William Woodall

0.23.1 (2022-04-08)

0.23.0 (2022-03-30)

0.22.0 (2022-03-28)

0.21.1 (2022-03-01)

• Added case for instances of ExecuteLocal in resolveProcess function (#587)
• Add compatitibility with pytest 7 (#592)
• Contributors: Matt Lanting, Shane Loretz

0.21.0 (2022-01-14)

• Renamed three files from example_processes (#573)
• Fix launch_testing README.md proc keyword to process. (#554) (#560)
• Contributors: Jacob Perron, Khush Jain

0.20.0 (2021-11-29)

• Declare frontend group dependency & use explicit dependencies in launch_testing (#520)
• Update maintainers to Aditya Pande and Michel Hidalgo (#559)
• Updated maintainers (#555)
• First prototype of native pytest plugin for launch based tests (#528)
• Adding Executable description class (#454)
• Add a "hello world" style example (#532)
• Contributors: Aditya Pande, Audrow Nash, Christophe Bedard, Ivan Santiago Paunovic, roger-strain

0.19.0 (2021-07-15)

0.18.0 (2021-06-18)

0.17.0 (2021-04-06)

0.16.0 (2021-03-19)

• Use unittest.mock instead of mock (#487)
• Contributors: Michel Hidalgo

0.15.0 (2021-01-25)

0.14.0 (2020-12-08)

0.13.0 (2020-11-04)

• Update package maintainers (#465)
• Contributors: Michel Hidalgo

0.12.0 (2020-08-18)

0.11.1 (2020-08-14)

0.11.0 (2020-08-04)

• Disable cleanup of test cases once they have been run (#406)
• Fix max() with empty sequence (#440)
• Use unittest.TestCase.id() for pytest failure reprs. (#436)
• Use unittest.TestCase.id() to put together jUnit XML output. (#435)
• Claim ownership (#433)
• Contributors: Dirk Thomas, Michel Hidalgo, Scott K Logan, William Woodall

0.10.2 (2020-05-26)

• Set junit_family to xunit2 in pytest.ini
• Stop using implicit variables in example testing.
• Switch to from_parent to remove deprecation warning.
• Fix new flake8 errors. (#420)
• Remove uses of deprecated ready_fn. (#419)
• Contributors: Chris Lalancette, Michel Hidalgo

0.10.1 (2020-05-08)

• fixed depcrecation warning of imp to importlib (issue #387) (#407)
• wait_for_ouput() repr includes actual text (#408)
• Contributors: Shane Loretz, Zahi Kakish

0.10.0 (2020-04-24)

• Improve jUnit output for launch tests when run with py.test (#404)
• avoid deprecation warning, use from_parent (#402)
• Warn that old-style ready_fn and test attributes will be deprecated (#346)
• more verbose test_flake8 error messages (same as ros2/launch_ros#135)
• stop using constructors deprecated in pytest 5.4 (#391)
• Add the ability to assert in stdout or stderr. (#378)
• Add delay parameter to retry_on_failure decorator (#390)
• Make RegisterEventHandler describe its sub-entities (#386)
• Import test file without contaminating sys.modules (#360)
• Update reference to example launch test file (#363)
• Use imperative mood in docstrings. (#362)
• Fix a documentation typo. (#361)
• Fix junit XML when launch dies early (#358)
• Contributors: Chris Lalancette, Dan Rose, Dirk Thomas, Jacob Perron, Michel Hidalgo, Peter Baughman, Steven! Ragnarök

0.9.5 (2019-11-13)

• Make launch_testing.markers.retry_on_failure decorator more robust. (#352)
• Contributors: Michel Hidalgo

0.9.4 (2019-11-08)

• Fix a small typo in the launch_testing README. (#351)
• Contributors: Chris Lalancette

0.9.3 (2019-10-23)

0.9.2 (2019-10-23)

• Support launch test reruns when using pytest (#348)
• Support CLI commands testing (#279)
• Contributors: Michel Hidalgo

0.9.1 (2019-09-25)

• Optionally remove ready fn arg from generate_test_description (#322)
• Contributors: Michel Hidalgo, Peter Baughman

0.9.0 (2019-09-18)

• install package manifest (#330)
• Unindent setup.cfg options. (#326)
• Use renamed remove_ansi_escape_sequences. (#302)0
• Enable launch test discovery in pytest (#312)
• Support LaunchService injection into pre-shutdown tests. (#308)
• Add assertWaitForStartup method to match assertWaitForShutdown (#278)
• Fix a simple typo in an error message. (#301)
• Fix launch_testing output filtering (#296)
• Revert "Revert "[execute_process] emulate_tty configurable and defaults to true"" (#277)
• Fix formatting (#262)
• Fix proc lookup for processes with multiple command-line arguments (#229)
• Remove ros domain ID dependency (#256)
• Contributors: Chris Lalancette, Dirk Thomas, Esteve Fernandez, Michel Hidalgo, Peter Baughman, William Woodall, ivanpauno

0.8.3 (2019-05-29)

• Changed behavior to use --isolated if no ROS_DOMAIN_ID is set to help parallel testing. (#251)
• Contributors: Peter Baughman

0.8.2 (2019-05-20)

• add non-asserting waitFor method (#243)
• Enable reuse of launch testing functionality (#236)
• Stop randomizing ROS_DOMAIN_ID by default in launch tests (#240)
• Contributors: Dirk Thomas, Michel Hidalgo

0.8.1 (2019-05-08)

0.8.0 (2019-04-13)

• Added test actions. (#178)
• Fixed test_env_testing test (#200)
• Dropped legacy launch package. (#191)
• Migrated legacy launch API tests. (#167)
• Contributors: Dirk Thomas, Michel Hidalgo, ivanpauno

0.7.3 (2018-12-13)

0.7.2 (2018-12-06)

0.7.1 (2018-11-16)

• Fixed setup.py versions (#155)
• Contributors: Steven! Ragnarök

0.7.0 (2018-11-16)

• Fixed lint warnings from invalid escape sequences (#151) Use raw strings for regex patterns to avoid warnings.
• Fixed linter errors from #131. (#132)
• Added class to provide some limitted testing options (#131)
• Moved launch_testing into launch_testing.legacy namespace (#130)
• Contributors: Dirk Thomas, Jacob Perron, Steven! Ragnarök

0.6.0 (2018-08-20)

0.5.2 (2018-07-17)

0.5.1 (2018-06-27)

0.5.0 (2018-06-19)

• Updated to use new launch.legacy namespace (#73)
• Contributors: Dirk Thomas, Mikael Arguedas, William Woodall