timgdavies
commented
6 years ago Yes! This is something we really want to work on.
We've been considering a couple of options here including:
- Trying to capture a RegEx for each registry as part of our meta-data;
- Creating a framework for custom functions for each registry that can help with validation, normalisation and resolution of identifiers;
There is a research challenge of identifying all the possible identifiers that can exist in a given series - as there are often edge cases not immediately obvious from research (e.g. legacy identifiers, special types of entity) so we need to take care to avoid a 'computer says no' system that upsets implementations.
Sergii Pavliuk, TI Ukraine at PDF CEE asked whether org-id.guide could provide the identifier format / validation rules for each registry. That information could be used by data entry systems to improve the quality of entered data, and could be used by data users to validate the published data.