fiberassign input for newexp-desi about desisim

[sbailey]

Just had a chat with Lado and Jaime; updating the ticket for the record. We'll start with a tool for generating spectra for a brick from an ra,dec,z,truetype truth file, and then separately update newexp-desi to use that tool and/or the pre-cached results from that tool.

The catalog truth file format is documented at http://desidatamodel.readthedocs.io/en/latest/DESI_TARGET/truth.html

The output "simbrick" (?) file with spectra per brick would be similar to the FLUX and METADATA HDUs of the simspec files described at http://desidatamodel.readthedocs.io/en/latest/DESI_SPECTRO_SIM/PIXPROD/NIGHT/simspec-EXPID.html . They should not include the sky or photon-level information that the simspec files have. i.e. simbrick files just describe the true objects, while the simspec files contain additional information about a particular observation of those objects.