ISO 3166-2 apostophe inconsistencies

In ISO 3166-2 (and perhaps other formats too) different apostrophes are used. This is okay as long as they don't vary within regions of a country. See e.g. ‘Ajmān (first character) and Ra’s al Khaymah (third character) for Emirate. Perhaps some automated detection can be useful and subsequent upstream reporting?