Refactor

[boulund]

I took the liberty to go through the code and here suggest some structural improvements along with some minor refactors to make the code easier to read overall and potentially extend in the future.

Further future changes that I would recommend include:

• Packaging as a Python package, for easy installation with pip, etc.
• Packaging as a conda package
• Using Python's logging module for info, warnings, and error messages instead of simply printing to stdout
• Improve exception handling: the except clauses for the pd.read_csv calls for reading both VALENCIA and metadata input currently catches everything which is not considered best practice
• Create a requirements.txt file to specify the minimum required version of Pandas needed to run VALODY (I think VALODY should work as intended on even very old versions of Pandas)

[luhugerth]

When the user does not input all necessary CST, the error message gives them in a random order like:

the following CSTs must be included: {'III-A', 'IV-B', 'III-B', 'IV-A', 'IV-C4', 'IV-C3', 'IV-C2', 'II-A', 'IV-C1', 'IV-C0', 'II-B', 'I-A', 'V', 'I-B'}

is it possible to make them come out alpha-numerically?

[boulund]

The set data structure is naturally unordered, but of course it could be converted to a list and sorted before printing the error message if you want to, e.g.

            print(f"ERROR: When using subtypes, the following CSTs must be included: {sorted(ALL_SUBTYPE_CSTs)}")