launch_pal

Utilities for simplifying some common ROS2 launch operations.

get_pal_configuration

Implementation of the PAL's PAPS-007 standard for configuration management.

Retrieves all the parameters, remappings and arguments for a given node by looking for ament_index-registered YAML configurations file. It properly handle overloading of parameters, enabling for instance to have a default configuration and a specific configuration for a given robot family or robot unit.

User overrides

Users can provide local overrides via configuration files in $HOME/.pal/config.

For instance, creating a file ~/.pal/config/default_volume.yml with the content:

/volume:
  ros__parameters:
    default_volume: 75

would override the ROS parameter default_volume for the node /volume.

This is useful for eg persist user configuration across robot reboots.

The default location of user configuration is $HOME/.pal/config. It can by changed by setting the environment variable $PAL_USER_PARAMETERS_PATH.

Automatic command-line arguments

If get_pal_configuration is called with cmdline_args=True (default), it will automatically add command-line launch arguments for all parameters in the configuration file. This allows for easy configuration of the node via the command line (see for instance --show-args).

Alternatively, the user can provide a list of parameters to be exposed as command-line arguments. This is useful for instance to expose only a subset of parameters.

This behavior can be disabled by setting cmdline_args=False.

Usage

#...
from launch_pal import get_pal_configuration

def generate_launch_description():

    ld = LaunchDescription()

    config = get_pal_configuration(pkg='pkg_name',
                                   node='node_name', 
                                   ld=ld, # optional if cmdline_args = False
                                   cmdline_args=[Bool|list]) # optional, True by default
                                   )
    my_node = Node(
        name='node_name',
        namespace='',
        package='pkg_name',
        executable='node_executable',
        parameters=config['parameters'],
        remappings=config['remappings'],
        arguments=config['arguments'],
    )

    # ...

    ld.add_action(my_node)

    return ld

robot_arguments

Contains classes to read launch argument settings directly from a YAML file, grouped per robot type. For each argument the name, the description, default value and possible choices are provided. The classes can be imported to remove boilerplate of robot launch arguments.

One special class, RobotArgs, contains all available launch arguments for PAL Robots. These arguments consist of only a name and description.

Example:

from launch_pal.arg_utils import LaunchArgumentsBase
from launch_pal.robot_arguments import TiagoArgs, RobotArgs
from launch.actions import DeclareLaunchArgument
from dataclasses import dataclass

@dataclass(frozen=True)
class LaunchArguments(LaunchArgumentsBase):

    # Tiago specific settings
    wheel_model: DeclareLaunchArgument = TiagoArgs.wheel_model
    # Robot agnostic argument
    base_type: DeclareLaunchArgument = RobotArgs.base_type

arg_utils

Contains utilities for declaring launch arguments and removing boiler plate.

An example of the use can be found in the launch_tutorial package that compares the standard boilerplate with the an updated version of the launch file. This updated version also provides a structured architecture that seperates launch arguments from other launch actions.

LaunchArgumentsBase: A dataclass that contains only DeclareLaunchArgument objects. The class is used to ease the process of adding launch arguments to the launch description. Has member function add_to_launch_description to automatically add all launch arguments to the launch description.

read_launch_argument: Used in Opaque functions to read the value of a launch argument and substitute it to text.

param_utils

Contains utilities for merging yaml parameter files or replace parametric variables in a param file.

parse_parametric_yaml: Checks yaml files for variables of layout ${VAR_NAME} and parses them. Parsing is done by giving a dictionary as input:

parse_dict = { VAR_NAME_1: value_1,
               VAR_NAME_2: value_2}

merge_param_files: Merges multiple yaml files into one single file to be loaded by a node.

include_utils

Contains utilities to reduce the boilerplate necessary for including files.

include_launch_py_description: Include a python launch file.

include_scoped_launch_py_description: Include a python launch file but avoid all launch arguments to be passed on by default. Any required launch arguments have to explicitly passed on to the launch file.

    scoped_launch_file = include_scoped_launch_py_description(pkg_name='my_pkg', 
    paths=['launch','my_file.launch.py'],
    launch_arguments={ 'arg_a': DeclareLaunchArgument('arg_a'),
                       'arg_2': DeclareLaunchArgument('arg_b'),
                       'arg_c': LaunchConfiguration('arg_c'),
                       'arg_d': "some_value' }
    env_vars=[SetEnvironmentVariable("VAR_NAME", 'value)]
    condition=IfCondition(LaunchConfiguration('arg_a')))

NOTE: This mimics the behavior of including launch files in ROS 1. Helpful in large launch files structures to avoid launch arguments to be overwritten by accident.

composition_utils

Contains utilities to reduce the boilerplate necessary for using ROS 2 components

generate_component_list: generates a list of composable nodes from a YAML and a package name, ready to be added or loaded into a ComposableNodeContainer:

components:
  <COMPONENT-NAME>:
    type: <DIAGNOSTIC-TYPE>
      ros__parameters: 
        name: <FULL-DIAGNOSTIC_DESCRIPTOR>
        [<REST-OF-PARAMS>]

It can be used from a launch file like:

component_list = generate_component_list(components_yaml, pkg_name)

And then added normally to a container:

container = ComposableNodeContainer(
    name="container_name",
    namespace="",
    package="rclcpp_components",
    executable="component_container",
    composable_node_descriptions=component_list,
)

Actions

CheckPublicSim

Raises an exception if the is_public_sim argument is being used correctly, that is, ensure that when using a simulation outside PAL the argument is set to true.

ValidateLaunchArgs

Checks that all the passed arguments using ros2 launch are declared in the launch file. This prevents passing uncorrectly typed arguments to a launch file, which would result in unexpected behaviours as the defaults would be used without warning.

validate_launch_args = ValidateLaunchArgs(launch_args=launch_args)
launch_description.add_action(validate_launch_args)

ValidateXacroArgs

It does two things:

• Checks that all the arguments that are being passed from the launch file to the xacro are declared in it, failing if not.
• Checks that all the declared arguments in the xacro are receiving a value from the launch file, giving a warning if not. This allows to know exactly which argument is being used in the robot description.

validate_xacro_args = ValidateXacroArgs(xacro_path=xacro_file_path, xacro_input_args=xacro_input_args)
launch_description.add_action(validate_xacro_args)

robot_utils (DEPRECATED)

Declare a single launch argument given by the robot name.

Example:

robot_name = 'tiago'
laser_model_arg = get_laser_model(robot_name)