Changed ndc_parser to create separate CS/MOA lists

[VidhiKulkarni]

The code reads the data files and writes the unique CS and MOA values from the pharm_classes into a list. The lists are in two separate files for easier viewing. I created a new function in the ndc_parser.py file called custom_dict_sweep, as the original dict_sweep function was causing errors because of unwanted values in the dictionary. The custom_dict_sweep function removes the empty lists/dicts from the dictionary (it cleans up the dictionary by removing unwanted fields from the data). I also added the print_list function in the ndc_parser.py file to print the values in two separate files.