ladewig
commented
1 year ago I dug into this a bit more, and I figured out that the default non-strict regex matches both the UK and Falkland Islands. But when you use "parse", it only returns the first match (UK). If you use _parse_registration (which parse calls), it returns both. Both results have the same priority of 0 in this example. These both fail for strict because the regex requires a second hyphen which is not correct for these entries. removing the "-{0,1}" from the regex for these (as well as others that don't use hyphens: US, Japan, South Korea, and Dominican Republic) would fix the issue between strict/non-strict.
For those entities that have their own registries that are using a range carved out of a parent allocation, such as the Falkland Islands and others under the UK, one solution would be to adjust priorities to prioritize them over the parent and to also include their ICAO 24-bit ranges. It appears this was done for some French entries, such as Tahiti, giving them a priority of 2. The same could be done for these UK entities.
Checking a Falkland Islands registration like "VP-FMC, returns the United Kingdom rather than the Falkland Islands. Given that their is a specific rule for the Falklands, I assume that's not the expected behavior.
There's also a typo. The entries say "Falk Islands" rather than "Falkland Islands".