Allow integers in conversion to float array messages

[jtbandes]

Public API Changes
None

Description
Fixes #764 by using ndarray/array assignment operations that accept ints as well as floats, rather than directly assigning the incoming list value, which failed at the msg/idl generated python code's validation step.

Before this change, the newly added tests would fail with this error for dynamic-length arrays:

2: ___________________ TestMessageConversion.test_float32array ____________________
2: 
2: self = <test.internal.test_message_conversion.TestMessageConversion testMethod=test_float32array>
2: 
2:     def test_float32array(self):
2:         def test_float32_msg(rostype, data):
2:             msg = {"data": data}
2:             inst = ros_loader.get_message_instance(rostype)
2:             c.populate_instance(msg, inst)
2:             self.validate_instance(inst)
2:             return inst.data
2:     
2:         for msgtype in ["TestFloat32Array"]:
2:             rostype = "rosbridge_test_msgs/" + msgtype
2:     
2:             # From List[float]
2:             floats = list(map(float, range(0, 256)))
2:             ret = test_float32_msg(rostype, floats)
2:             np.testing.assert_array_equal(ret, np.array(floats))
2:     
2:             # From List[int]
2:             ints = list(map(int, range(0, 256)))
2: >           ret = test_float32_msg(rostype, ints)
2: 
2: ../../rosbridge_library/test/internal/test_message_conversion.py:304: 
2: _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 
2: ../../rosbridge_library/test/internal/test_message_conversion.py:290: in test_float32_msg
2:     c.populate_instance(msg, inst)
2: ../../install/rosbridge_library/lib/python3.8/site-packages/rosbridge_library/internal/message_conversion.py:177: in populate_instance
2:     return _to_inst(msg, inst_type, inst_type, inst)
2: ../../install/rosbridge_library/lib/python3.8/site-packages/rosbridge_library/internal/message_conversion.py:294: in _to_inst
2:     return _to_object_inst(msg, rostype, roottype, inst, stack)
2: ../../install/rosbridge_library/lib/python3.8/site-packages/rosbridge_library/internal/message_conversion.py:403: in _to_object_inst
2:     setattr(inst, field_name, field_value)
2: _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 
2: 
2: self = rosbridge_test_msgs.msg.TestFloat32Array(data=[0.0, 1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 7.0, 8.0, 9.0, 10.0, 11.0, 12.0, 13....239.0, 240.0, 241.0, 242.0, 243.0, 244.0, 245.0, 246.0, 247.0, 248.0, 249.0, 250.0, 251.0, 252.0, 253.0, 254.0, 255.0])
2: value = [0, 1, 2, 3, 4, 5, ...]
2: 
2:     @data.setter
2:     def data(self, value):
2:         if isinstance(value, array.array):
2:             assert value.typecode == 'f', \
2:                 "The 'data' array.array() must have the type code of 'f'"
2:             self._data = value
2:             return
2:         if __debug__:
2:             from collections.abc import Sequence
2:             from collections.abc import Set
2:             from collections import UserList
2:             from collections import UserString
2: >           assert \
2:                 ((isinstance(value, Sequence) or
2:                   isinstance(value, Set) or
2:                   isinstance(value, UserList)) and
2:                  not isinstance(value, str) and
2:                  not isinstance(value, UserString) and
2:                  all(isinstance(v, float) for v in value) and
2:                  True), \
2:                 "The 'data' field must be a set or sequence and each value of type 'float'"
2: E           AssertionError: The 'data' field must be a set or sequence and each value of type 'float'
2: 
2: ../../install/rosbridge_test_msgs/lib/python3.8/site-packages/rosbridge_test_msgs/msg/_test_float32_array.py:131: AssertionError

And this error for fixed-length arrays:

2: self = rosbridge_test_msgs.msg.TestFloat32BoundedArray(data=array([ 0.,  1.,  2.,  3.,  4.,  5.,  6.,  7.,  8.,  9., 10., 11., 12.,
2:        13., 14., 15.], dtype=float32))
2: value = [0, 1, 2, 3, 4, 5, ...]
2: 
2:     @data.setter
2:     def data(self, value):
2:         if isinstance(value, numpy.ndarray):
2:             assert value.dtype == numpy.float32, \
2:                 "The 'data' numpy.ndarray() must have the dtype of 'numpy.float32'"
2:             assert value.size == 16, \
2:                 "The 'data' numpy.ndarray() must have a size of 16"
2:             self._data = value
2:             return
2:         if __debug__:
2:             from collections.abc import Sequence
2:             from collections.abc import Set
2:             from collections import UserList
2:             from collections import UserString
2: >           assert \
2:                 ((isinstance(value, Sequence) or
2:                   isinstance(value, Set) or
2:                   isinstance(value, UserList)) and
2:                  not isinstance(value, str) and
2:                  not isinstance(value, UserString) and
2:                  len(value) == 16 and
2:                  all(isinstance(v, float) for v in value) and
2:                  True), \
2:                 "The 'data' field must be a set or sequence with length 16 and each value of type 'float'"
2: E           AssertionError: The 'data' field must be a set or sequence with length 16 and each value of type 'float'