Cannot use random value as list index in a constraint

[felixdube]

I am wondering if it is possible to use a random value as a list index inside a constraint. It does not seem to work as I expect. Here is a minimal example, which produce the following error:
TypeError: list indices must be integers or slices, not rand_uint32_t

import vsc

@vsc.randobj
class my_item():
    def __init__(self):
        self.a = vsc.rand_uint32_t()
        self.b = vsc.rand_uint32_t()
        self.c = [2,4,6,8]        

    @vsc.constraint
    def ab_c(self):
        vsc.solve_order(self.a, self.b)
        self.a < len(self.c)
        self.b < self.c[self.a]	

item = my_item()

I have found a workaround, but it might not be very efficient depending on the size of the list.

    @vsc.constraint
    def ab_c(self):
        vsc.solve_order(self.a, self.b)
        self.a < len(self.c)
        for i in range(len(self.c)):
            with vsc.if_then(self.a == i):
                self.b < self.c[i]

[mballance]

@felixdube, interestingly enough, your workaround is more or less exactly what would happen "under the hood" to implement the self.b < self.c[self.a] constraint.
Any feedback on how common you see this case to be in your work? If it's common (ie a workaround will be a real pain), that certainly will impact the priority of addressing it sooner vs later.

[felixdube]

@mballance, thank you for your feedback! I can live with the workaround for now. I'll comeback if ever it becomes a real pain.