JSON Schema files defining three OpenActive data profiles:
- Core (powering activity finders)
- Social Prescribing
- Accessibiity
Note that, despite their considerable commonalities, the various profiles do not cross-reference each other. This is to preserve the possibility of divergence between the profiles (as seen, for instance, with the different profiles of activity
between the Social Prescribing and Core profiles). The profiles should thus be treated as self-contained and independent, with no references linking them to each other.
- most constraints right now are only on string-length. These should be tightened into regexes where possible
- Add 'description' documentation to the Schema files
- Think about the implications of the different profiles for different processing chains. Should the normaliser ideally produce a superset valid against all profiles?