Use parser for beam density

[MaxThevenet]

This PR adds the option to parse the beam density, for the fixed-ppc option. To be used with caution as the fixed-ppc option may result in impractically large number of beam particles, e.g. several particles per cell in the whole simulation box. To avoid this, please make sure you limit the creation of beam particles to relevant regions, e.g.

my_constants.w = 10.e-6
my_constants.nb = 5.e23

beam.injection_type = fixed_ppc
beam.profile = parsed
beam.density(x,y,z) = nb * exp(-(x**2+z**2+y**2)/w**2)
beam.min_density = nb/100 # do not allocate beam particles where the density is <1% of the peak density
beam.radius = 200.e-6 # limit the region where beam particles are allocated, both transversely...
beam.zmin = -200.e-6 # ... and longitudinally
beam.zmax =  200.e-6

[AlexanderSinn]

At the very least the location of m_parser NEEDS to be changed so it doesn’t get transferred to the GPU.

[MaxThevenet]

Thanks! I addressed your comments in the last commit.