ROSwrapper is a wrapper for ROS2's rclpy. ROSwrapper uses object oriented programming to turn nodes into clean, self-contained objects that can be added or removed from the NodeControl class. Nodes are spun and stopped through use of a controller, as defined in the NodeControl class.

To create a publish node with a timer in rclpy:

def timer_callback():
    msg.data = 'Hello World'
    publisher.publish(msg)


rclpy.init()
node = rclpy.create_node('publisher')
publisher = node.create_publisher(String, 'topic')

msg = String()
timer_period = 0.5  # seconds
timer = node.create_timer(timer_period, timer_callback)

rclpy.spin(node)

node.destroy_timer(timer)
node.destroy_node()
rclpy.shutdown()

You can spin up a publisher in 13 lines of code using just rclpy. Additional publishers/subscribers are several lines each.

To publish the same thing with ROSwrapper:

nc = NodeControl()
nc.addnode(RosNode(name='publisher',
                   pub_data_type=String,
                   pub_chan='topic',
                   pub_rate=2,  # in Hz
                   pub_data='Hello World'))
nc.run()

You can spin up a publisher in 3 lines of code using ROSwrapper. Additional publishers/subscribers are 1 line of code each. Nodes are added to the NodeControl so everything can be cleanly executed and cleaned up.

The core nodes are defined in the base class RosNode. RosNode was intended to be derived in the event the user wishes to see specific behavior.

• An example of it running with the base node type can be found in example_basenode.
• An example of a derived node class can be found in the example_derivednode file.

If a user has multiple subscribers you can use a switch statement to interpret the message.

For more advanced usage you coud derive multiple classes from RosNode to handle each case. For example a robot may have IR sensors and wheels, each of which could have a custom class to deal with the information.

This still has quite a bit of work to be done before it's ready.

I was tasked with coding a robot to navigate a course while utilizing ROS2. I prefer object-oriented programming with working with something such as nodes so I built this to wrap rclpy into a more user-friendly product.