URL parser

https://user:pass@shop.example.co.uk:8443/products/shoes.html?color=blue&size=42#reviews
└─┬─┘   └───┬───┘ └───────┬────────┘ └┬─┘└────────┬─────────┘ └───────┬────────┘ └──┬──┘
scheme  userinfo        host        port        path                query       fragment

The host name splits further. In shop.example.co.uk, co.uk is the public suffix, example.co.uk is the domain someone registered and shop is a subdomain. Which suffixes have two parts is not a rule you can compute; it comes from the Public Suffix List that browsers ship. The parser knows the common ones.

Besides the parts, the tool points out things that often cause bugs:

Values in the table are shown decoded. To decode or encode a single value by hand, use the URL decoder and encoder.

JavaScript (browser and Node.js)

const u = new URL('https://shop.example.com:8443/p/shoes?size=42&size=43#reviews');
u.hostname;                  // 'shop.example.com'
u.port;                      // '8443'
u.searchParams.get('size');  // '42'
u.searchParams.getAll('size'); // ['42', '43']

// Edit and rebuild
u.searchParams.delete('size');
u.searchParams.set('utm_source', 'newsletter');
u.toString(); // 'https://shop.example.com:8443/p/shoes?utm_source=newsletter#reviews'

URL.canParse('not a url');   // false, without try/catch

Python

from urllib.parse import urlsplit, parse_qs, urlencode, urlunsplit

u = urlsplit('https://shop.example.com:8443/p/shoes?size=42&size=43#reviews')
u.hostname, u.port, u.path   # ('shop.example.com', 8443, '/p/shoes')
parse_qs(u.query)            # {'size': ['42', '43']}

q = urlencode({'utm_source': 'newsletter'})
urlunsplit(u._replace(query=q))

PHP

$u = parse_url('https://shop.example.com:8443/p/shoes?size=42#reviews');
$u['host'];   // 'shop.example.com'
$u['port'];   // 8443
parse_str($u['query'], $params); // $params['size'] === '42'

Go

u, err := url.Parse("https://shop.example.com:8443/p/shoes?size=42&size=43#reviews")
u.Hostname()            // "shop.example.com"
u.Port()                // "8443"
u.Query()["size"]       // []string{"42", "43"}
q := u.Query(); q.Set("page", "2"); u.RawQuery = q.Encode()

Relative URLs need a base. new URL('/p/shoes') throws; pass a base as the second argument: new URL('/p/shoes', 'https://example.com'). This tool adds https:// when the scheme is missing.

Do not parse URLs with a regex or split(). A ? inside the fragment, an @ in user info, an IPv6 host in brackets or a # before the query all break simple patterns. Parsers that disagree with the browser are a known source of SSRF and open-redirect bugs, so check redirect targets with new URL() and compare the origin.

Encoding changes on rebuild. URLSearchParams writes spaces as + and re-escapes every value, so q=red%20shoes&x=a!b comes back as q=red+shoes&x=a%21b after any edit. This tool keeps parameters you did not edit exactly as they were, and only re-encodes the ones you changed.

Building links for a site? The meta tag generator previews how a URL looks when shared, and the curl converter turns a request into code. For API clients and testing tools, see API tools on DevHunt.