Neural Operators Meet Energy-based Theory: Operator Learning for Hamiltonian and Dissipative PDEs