This MR provides a new MaranaX class, which is just a subclassed hRIXS class to enable parallelized centroiding. The usage will still be similar to the current implementation.
An example usage and some benchmarks can be found https://max-jhub.desy.de/hub/user-redirect/lab/tree/GPFS/exfel/data/scratch/carinanc/SCS/hRIXS/240529_MaranaX-class.ipynb. One can see that tenfold performance gain, hooray!
TODO: add a changelog :V (also for !287 (merged) that I forgot)