url-normalize

Turn any URL-like string into a clean https:// URL, or null.

url-normalize turns any URL-like string into a clean, canonical https:// URL, and returns null if it can't. By default it uses HTTPS, removes www, sorts query params and strips default ports. It handles bare domains, protocol-relative URLs and Unicode to Punycode in both directions. It's 980 bytes with brotli, has one dependency and ships CJS and ESM.

It's for JavaScript and TypeScript code that takes URLs from people or scraped pages and needs to compare, store or show them. Invalid input gives null, so there's no try/catch. extractDomain and humanizeUrl come with the package.

There's also normalize-url. The differences are dual CJS and ESM exports, null instead of throwing, and the built-in domain and display utilities.

Install

npm i url-normalize

import { urlNormalize } from "url-normalize"

urlNormalize("example.com")
// → "https://example.com"

urlNormalize("//www.example.com:443/../foo/bar?b=2&a=1#tag")
// → "https://example.com/foo/bar?a=1&b=2"

urlNormalize("example") // → null