Make code less fragile and more extendable - Use ASE!

[lucydot]

At the moment the parsing of the FHI-aims output files is done via regex. The upside to this is that it reduces our need for dependencies (using the re core Python library only). The downside is that this makes our parsing fragile to changes made to the FHI-aims output file format. Instead, we could parse calculation data using ASE. ASE has a large user base and so is more likely to be updated when changes occur! This will also allow easier extension to parsing the other DFT codes supported by ASE.