string().hostiname() - validation of non-ASCII chars containing domains

[viktorija]

Hi,

There's an issue with how Joi.string().hostname() validates domain names with non-ASCII characters.
E.g. domain name: cptømrerfirma.dk fails validation, although it's a valid and existing domain.

Thanks,
Viktorija

[Marsup]

As is expected in the RFC. It was already discussed in #1298, but it seems no one was interested in taking up this task.

[JoaoManoel]

Hi.

Can I get this issue?

I'm thinking about changing the hostRegex to:

hostRegex: /^(([a-zA-ZÀ-ÿ0-9]|[a-zA-ZÀ-ÿ0-9][a-zA-ZÀ-ÿ0-9\-]*[a-zA-ZÀ-ÿ0-9])\.)*([A-Za-zÀ-ÿ0-9]|[A-Za-zÀ-ÿ0-9][A-Za-zÀ-ÿ0-9\-]*[A-Za-zÀ-ÿ0-9])$/

What do you think?

[hueniverse]

This is not as simple as changing the regex. I need to consolidate a few places that deal with domain names first.

[BanovMiroslav]

I wonder if this was really correct? Normally internationalized domains are punycode encoded.
So cptømrerfirma.dk is encoded as xn--cptmrerfirma-xjb.dk

So with that in mind should the validation of hostnames be accepting non-ASCII?