But there would not be anything like mirror or any object that will reflect the light from its own straight way. Here light means any visible lights. "VIBGYOR" any one.
No! In vacuum it is impossible. In particular statement "light is electromagnetic wave therefore it can be deviated from its straight path by applying highly non uniform magnetic field" is wrong.
Maxwell equations in vacuum are linear. Two solutions of maxwell equations as for instance a propagating light wave and a static magnetic field cam bu summed together and it will be a solution of the Maxwell equations as well. But the simple sum o fields does not introduce any bending or interactions.
Instead, in free space interference can lead to bending like behavior. Look at "Accelerating Airy beams". But it is just an interference of plane waves. I am attaching two interesting papers.
Yes accelerating airy beams travel in curves. It is possible to design diffractive optical elements or holographic optical elements with suitable phase functions such that light passing through them travel in curves.
In vacuum no magnetic field will deviates the light (gravitational field yes....) but yes you can use any material which optical properties change with magnetic field: called magneto-optics effect. Like this : https://en.wikipedia.org/wiki/Magneto-optic_Kerr_effect
I basically believe that, this bending by magnetic field is possible in two specific cases. 1. if the magnetic field will be big enough can change the energy of the light beam and it would be happen. 2. if the light shows a particle behavior or in the case of photon position, this is because of changing energy of the photon and changing the electric field or velocity.
Light consist of particles called as photon which have no charge on them, which mean you can't bend light via electromagnetic fields. However you can bend them by gravitational field.
Compton scattering suggests that light can be bent in electromagnetic fields but that involves interacting with a charged particle. This question seems to be fundamental and require detailed knowledge of particles and interactions that can't be studied in detail as they are happening, but only before and after the interaction. If the photon has no charge, as currently believed, then it would not couple to an external B or E field. It is a propagating disturbance in a field that can collapse to a particle during various interactions. This question lies near the heart of wave-particle duality.