standard C float to custom floating point representation using 32-bit integer and bit-wise operators?

Question

I\'m trying to write a small program to convert a C float to a 12-bit floating point representation using a 32-bit integer and bit-wise operators.

The order goes: [unused