Hamilton-Jacobi Reachability in Reinforcement Learning: A Survey