Parse MS^n files

[benbowen]

Since @wholtz and I were just talking about the file converter and MS^n requirements, here is the code I wrote to parse these kinds of files. It works well, but doesn't make an output that lines up with our current hdf5 format in metatlas. The HDF5 file needs a few more attributes to store this kind of fragmentation data: the collection of spectra for convenience, the parent-child relationships, and perhaps additional metadata about each spectrum.

See the BLINK repo for how I parse MS^n files with pymzml:

This will be pretty common in a year or two; so getting some pieces in place is probably not a bad idea.