How to iterate over the circuit/pykernel

[ACE07-Sev]

Greetings there,

Hope all are well. I have a quick question I'd appreciate guidance on. I'm trying to see what operations I have added to a circuit, but can't seem to find any methods to do it.

import cudaq

qc = cudaq.make_kernel()
qr = qc.qalloc(2)
qc.h(qr[0])
qc.cx(qr[0], qr[1])

# HOW TO DO?
for gate in qc:
    print(gate)

[ACE07-Sev]

@1tnguyen Apologies for the ping. Can you kindly guide me on this? I need it to be able to add my own parser until/unless #3548 is done.

[1tnguyen]

As CUDA-Q kernel contains the whole MLIR tree (in a couple of dialects, such as Quake, CC, etc.), it's not possible to iterate over it in the linear gate operation sequence shown above.

To get just the list of quantum operation, we need to 'trace' the kernel.
An example of similar use case is draw (https://github.com/NVIDIA/cuda-quantum/blob/main/runtime/cudaq/algorithms/draw.h)

[ACE07-Sev]

Is there a python facing function for this? So that I can pass the PyKernel and get the list.