extract_from_url should handle url without protocol-scheme

[nsteinberg-r7]

How to reproduce:
call extract_from_url with //mail.google.com/mail as input

result will be Invalid Domain Error
expected behavior is to handle the case of missing protocol and return {subdomain: mail, domain: google, suffix: com}

[wavenator]

Technically this is not a valid URI but a URI reference. https://en.wikipedia.org/wiki/Uniform_Resource_Identifier#URI_references

We can support it, but not without a // at the beginning to distinguish between a valid and invalid URIs

[vihaanmody1]

Tldextract can extract with schemes

[elliotwutingfeng]

It appears that extract_from_url("//mail.google.com/mail") now works.