Remark on the persistence of spatial analyticity for cubic nonlinear Schrödinger equation on the circle