QGL2 programs cannot use parameters of qubits

[ahelsing]

QGL2 treats a qubit as just a label. Any parameter of the qubit cannot be used in a QGL2 program. Specifically, BasicSequences/BlankingSweeps shows a QGL program that you cannot write in QGL2.

Is this something we should support? If so, how? If not, do all such programs need to be written in QGL1 and given stubs for use in QGL2 programs?

[ahelsing]

Add to QGL2 codebase some helper QGL1 functions (say in qgl2/qgl1_util.py):

As simple as

def getPulseParam(q: qreg, param):
    return q.pulse_params['param']

Or fancier & more use specific:

def getPulseOfSpecificLength(pulse, q: qreg, desiredLength) -> pulse:
    return pulse(q, length=desiredLength-q.pulse_params["length"])

[ahelsing]

See Decoupling.CPMG. Using that getPulseParam idea may not work in general, because the QGL2 compiler tries to use it to ensure types work for the math (eg doing foo - getPulseParams(q, 'length') causes QGL2 to say it can't do float - None. The 2nd solution seems to work; see CPMG.

[ahelsing]

Per discussion today, using parameters of a qubit may be rare, such that it's OK if you have to do this in QGL1. So may not need to fix this; the workaround used above may be plenty.