Custom unitary op cannot be loaded in a kernel using a parameter

Required prerequisites

• Consult the security policy. If reporting a security vulnerability, do not report the bug using this form. Use the process described in the policy to report the issue.
• Make sure you've read the documentation. Your issue may be addressed there.
• Search the issue tracker to verify that this hasn't already been reported. +1 or comment there if it has.
• If possible, make a PR with a failing test to give us a starting point to work on!

Describe the bug

Custom unitary op cannot be loaded in a kernel if the name is constructed from a parameter.
The following example gives an error:

RecursionError: maximum recursion depth exceeded while calling a Python object

Steps to reproduce the bug

def register_custom_operation(unitary_index: int, matrix: np.ndarray):
    cudaq.register_operation(f"U_{unitary_index}", matrix)

kernel, s, i = cudaq.make_kernel(cudaq.State, int)
qubits = kernel.qalloc(s)
kernel.__getattr__(f'U_{i}')(qubits[0], qubits[1], qubits[2])

state = np.zeros(8)
state[0] = 1
state = np.array(state, dtype=cudaq.complex())
states = [state]
for i in range(0,3):
    matrix = flip_qubit(np.identity(8), i)
    register_custom_operation(i, matrix)

    cudaq_state = cudaq.State.from_data(state)
    state = np.array(cudaq.get_state(kernel, cudaq_state, i))
    states.append(state)

for s in states:
    counts = cudaq.sample(kernel_helper, s)
    print(counts)

Expected behavior

We might consider creating a special API instead of getting the attribute to make this code work

Is this a regression? If it is, put the last known working version (or commit) here.

Not a regression

Environment

• CUDA Quantum version:
• Python version:
• C++ compiler:
• Operating system:

Suggestions

No response

Start by running the Python reproducer around cudaq.register_operation, cudaq.make_kernel, kernel.getattr, and cudaq.get_state to confirm the recursion error. Trace how the parameter-built operation name is resolved, then verify that the custom unitary loads and executes when the name depends on a kernel parameter.