LISPR: An Options Framework for Policy Reuse with Reinforcement Learning