refactor code to add an intermediate json file that would be used to generate a language specific header file

[mofosyne]

For now since we are only dealing with CBOR and CoAP, we don't need to worry too much.
But when we want to start adding support for other language besides C, we would want to create an intermediate json file that would contain the generated variable/enum names.

This is because the logic of converting the csv table names into variable/enum name would be the common factor between two different language header conversions.

This would however make it not a single python program operations, but rather two. First one to convert from IANA csv files to our internal json representation, then the second program would read it and then generate the language specific code (along with the relevant backward compatibility handling that would be needed for each project).