Shows that classes of exact solutions of the Dirac equation can be constructed out of combinations of gauge functions and action functions. A simple example is constructed by this procedure, showing how the Klein paradox can be very simply discussed.