Message conversion fails to convert list of integers to float32[]

[barulicm]

Description
When trying to populate a ROS message that includes a float32[] field, if the JSON data for that field is a list of integer values, the conversion raises an error instead of converting the integer values to floats.

This can become problematic when the library on the JS side decides to truncate float values, such as with this:

let joy_msg = new ROSLIB.Message({ axes: [ 0.0 ] });

joy_publisher.publish(joy_msg)

• Library Version: 1.2.0
• ROS Version: Humble
• Platform / OS: Ubuntu 22.04

Steps To Reproduce
python3 example.py

# example.py
from json import loads
from rosbridge_library.internal import message_conversion, ros_loader

msg_inst = ros_loader.get_message_instance('sensor_msgs/Joy')
msg_json = loads('{ "header": { "stamp": { "sec": 0, "nanosec": 0 }, "frame_id": "" }, "axes": [0], "buttons": [] }')
# this fails because it can't find an array of floats in [0]
message_conversion.populate_instance(msg_json, msg_inst)

Expected Behavior
The conversion function should be able to convert the array of "ints" into a float32 array.

Actual Behavior
The following error is raised:

Traceback (most recent call last):
  File "/home/matt/example.py", line 8, in <module>
    message_conversion.populate_instance(msg_json, msg_inst)
  File "/opt/ros/humble/local/lib/python3.10/dist-packages/rosbridge_library/internal/message_conversion.py", line 177, in populate_instance
    return _to_inst(msg, inst_type, inst_type, inst)
  File "/opt/ros/humble/local/lib/python3.10/dist-packages/rosbridge_library/internal/message_conversion.py", line 294, in _to_inst
    return _to_object_inst(msg, rostype, roottype, inst, stack)
  File "/opt/ros/humble/local/lib/python3.10/dist-packages/rosbridge_library/internal/message_conversion.py", line 399, in _to_object_inst
    setattr(inst, field_name, field_value)
  File "/opt/ros/humble/local/lib/python3.10/dist-packages/sensor_msgs/msg/_joy.py", line 165, in axes
    assert \
AssertionError: The 'axes' field must be a set or sequence and each value of type 'float' and each float in [-340282299999999994960115009090224128000.000000, 340282299999999994960115009090224128000.000000]