d3q Code

A few fixes to properly apply LO-TO splitting when doing thermal transport calculation (and more in general, to correctly appl LO-TO sploitting on the grid used to integrate the self energy)

Courtesy of Michele Simoncelli.

Fixed compilation when using recent git QE version, with upflib
Restructured calculation of frequency-dependt properties (selfenergy, spectral weight, dielectric constant) and tentative implementation of IR reflectivity (there is still a 2pi missing, or something and only cubic systems for the moment)

Fixed compilation with new QE including upflib, and a bug when compiling without MPI spport

optitmized grids fixed (although still a stupid idea, better way to optimize is in progress)

removed debug line

fix compatibility with new QE

format

fix

MgO example