Linter R3: Seeded Randomness

[Ellpeck]

Combination of the query functions from #1313 and #1315:

Q.calls("our set of random consumer functions")
  .with(Info.LastCall, { callName: 'our set of random producer AND consumer functions' })
  .with(Info.Arguments)
  
// ensure that all last calls are sufficiently linked to a previous producer function 
// OR linked to a previous consumer function that is sufficiently linked again

With "our set of random functions" resulting from a csv or similar data store of functions that produce and consume randomness:

function, package, type,    condition
set.seed, base,    produce, constant_argument
runif,    stats,   consume                   
isample,  ...,     both                       

Because loops and functions like isample can create an arbitrary amount of random numbers, it both produces and consumes randomness!