simplification of sigmf-schema.json

[gmabey]

I reworked the schema to not use "anyOf" sections, and it seems to work ok still.
However, I suspect that there's something I don't understand about JSON schema; this is my first exposure to it.
As I am using json-schema-validator, this change results in the error descriptions being intelligible ...

[gmabey]

I'm not github savvy enough to know how best to link this PR, since it builds on another PR ... https://github.com/gmabey/SigMF/tree/removed_anyOf

[Teque5]

According to this the keywords are anyOf, oneOf, allOf, not. Does removing this imply anyOf? Not sure.

I don't have a strong feeling on it, but are there cases where we may want to use one of these other keywords? Will removing anyOf preclude that?

When I initially wrote the schemas I used this tool if it's helpful.

[gmabey]

My best understanding (right now) is that anyOf is more important of a feature when enforcing things that are more complicated than the presence of properties, which AFAICT we indicates through the use of required and additionalProperties.
Please don't ask me what I'm alluding to when I say "things that are more complicated" ... :-D
I can say for sure that with the validators I've applied, the commit seems to still enforce the optional/required nature that we understand to be appropriate.