is_iri should allow UTF-8 characters

[gu-stav]

Currently is_iri fails on any characters, which don't match this regex:

valid-url/index.js

Line 28 in 8d1fc52

if (/[^a-z0-9\:\/\?\#\[\]\@\!\$\&\'\(\)\*\+\,\;\=\.\-\_\~\%]/i.test(value)) return;

Example:

t.ok(is_uri('http://localhost/ä'), 'http://localhost/ä');

As far as I understand the main difference between URI and IRI is:

IRIs extend URIs by using the Universal Character Set, where URIs were limited to ASCII, with far fewer characters.

https://en.wikipedia.org/wiki/Internationalized_Resource_Identifier

Therefore is_iri should allow all UTF-8 characters and these exports aren't correct:

module.exports.is_uri = is_iri;
module.exports.isUri = is_iri;

Instead there should be two exports: is_iri and is_uri.

I saw there wasn't any commit to this repo since 2015, but since this affects a Gatsby issue I was wondering, what you plans with the module are? Do you think about handing it over, @ogt?

The module is used heavily according to npms download numbers, so it might be in the interest of the community to give it some 💌 .

[adrienv1520]

Hi @gustavpursche,

I was facing the same issues with valid-url but also validator so I decided to build a module as reliable as possible strictly based on RFC-3986: https://github.com/adrienv1520/node-uri

The main features of this project are:

• parse any URI (URNs, URLs, URIs with IDNs support, etc.);
• get the safe Punycode ASCII or Unicode serialization of a domain;
• check an URI, HTTP/HTTPS/Sitemap URL, IP, domain is valid with clear checking errors;
• encode/decode an URI, HTTP/HTTPS/Sitemap URL.

I hope it could help you.