I am experimenting with implementing minimal LR(1) algorithms/techniques, but I find it very difficult to verify the correctness of what I've implemented.
Is there a fast algorithm to determine if two LR(1) action/goto tables accept the same language, given that they use the same grammar rules?
Or preferably a stronger (but easier to falsify) test, that two LR(1) action/goto tables will always produce the same derivation for any input.