Some of the official examples are considered invalid

[LinusU]

It seems like the following language tags, picked right from Appendix A of bcp47, isn't considered valid. Is this a bug?

[
  'i-enochian',
  'sl-rozaj-biske',
  'x-whatever',
  'qaa-Qaaa-QM-x-southern',
  'de-Qaaa',
  'sr-Latn-QM',
  'sr-Qaaa-RS'
]

[mattcg]

Hi Linus, thanks for reporting this.

The tag 'i-enouchian' is deprecated, so:

tags('i-enochian').errors()[0].message;

Will give you: 'The tag \'i-enochian\' is deprecated.'

There's a spec question here: should a deprecated grandfathered tag be considered valid or invalid?

For all the others, those tags are valid according to the spec, so I guess we need to make some fixes to the parser.

[sneko]

@mattcg I think at least it worths it to document not valid means "not a current valid tag", BUT it respects a format that has existed and that could still exist on internet, website, software (and shows an example of using both valid and deprecation checks?)... Otherwise without going through issues I could have let it in my code without knowing I was not considering things I thought valid (to be honest, I saw that with unit testing 😄 ).

A solution could be to have a .valid({ considerDeprecated: true }) so it would not bring any breaking change (for current tag but also subtags).