Presumably they just split the “domain” part out of the URL on // and the first / and feed that into getaddrinfo, with the OS and DNS doing the rest?

But I agree, it’s definitely neat :)