Refactor the loading by list of filenames

[Marshall McDonnell]

We have the capability to instead of list runs as Runs: <file numbers> to also say Filenamess: <list of filenames> in the main script.

Example:

{  `Filenamess`: [filename1, filename2, filename3] }

Yet, this needs some refactoring, the steps need to be:

• Put the following lines of code into a parse_config function: https://github.com/marshallmcdonnell/mantid_total_scattering/blob/master/mantid_total_scattering.py#L394-L514
• Refactor a little bit and and maybe move to a different module
• Add the Filenames ability
• Test

[Marshall McDonnell]

Embarrassingly, this does exists in a current form here: https://github.com/marshallmcdonnell/mantid_total_scattering/blob/master/mantid_total_scattering.py#L340-353

Yet, the parse_config-style function, refactor, and testing is still needed.

[Marshall McDonnell]

Another thought would be to add an InputDirectory option. Then, with the Instrument, Runs, and InputDirectory for the data files, it could construct the same thing as the manually input, verbose Filenames option. This would replicate behavior similar to GUDRUN (a positive feature).