Robust, configurable URL pattern matching, conforming to the algorithm used by Chrome and Firefox browser extensions.
- Native ESM import (browser, Deno):
import { matchPattern } from 'https://esm.sh/[email protected]'
- NPM module (Node):
npm i browser-extension-url-match
This library uses the native URL
constructor and Array#flatMap
. Polyfills may be required if you need to support Internet Explorer or Node.js <= 11.X.
A live demo is available.
The matchPattern
function takes a pattern or array of patterns as input and returns a valid or invalid Matcher
object:
- If all input patterns are valid,
matcher.valid
will betrue
. - If one or more input patterns are invalid,
matcher.valid
will befalse
, andmatcher.error
will contain a diagnostic error object.
Calling matcher.assertValid
asserts that the matcher is valid and throws an error at runtime if it isn’t.
By default, matchers use Chrome presets with strict URL matching.
import { matchPattern } from 'browser-extension-url-match'
const matcher = matchPattern('https://example.com/foo/*').assertValid()
matcher.match('https://example.com/foo/bar') // ⇒ true
matcher.match('https://example.com/bar/baz') // ⇒ false
const matcher2 = matchPattern([
'https://example.com/foo/*',
'https://example.com/bar/*',
]).assertValid()
matcher2.match('https://example.com/foo/bar') // ⇒ true
matcher2.match('https://example.com/bar/baz') // ⇒ true
const matcher3 = matchPattern('<all_urls>').assertValid()
matcher3.match('https://example.com/foo/bar') // ⇒ true
const invalidMatcher = matchPattern('htp://example.com/*')
invalidMatcher.valid // ⇒ false
invalidMatcher.error // ⇒ TypeError: Scheme "htp" is not supported
invalidMatcher.assertValid() // throws TypeError at runtime
If the input patterns are hard coded, calling assertValid
is a way of telling the TypeScript compiler that they’re assumed to be valid. However, if patterns are supplied from user input or other sources with unknown integrity, it’s usually better to check the valid
property, which allows TypeScript to correctly infer the type:
const matcherInput = form.querySelector<HTMLInputElement>('input#matcher')!
const checkBtn = form.querySelector<HTMLButtonlement>('button#check')!
checkBtn.addEventListener('click', () => {
const matcher = matchPattern(matcherInput.value)
// type narrowing via ternary operator
matcherInput.setCustomValidity(matcher.valid ? '' : matcher.error.message)
matcherInput.reportValidity()
// type narrowing via `if ... else`
if (matcher.valid) {
const url = prompt('Enter URL to match against')
alert(matcher.match(url ?? '') ? 'Matched!' : 'Unmatched')
} else {
console.error(matcher.error.message)
}
})
You can customize matchPattern
by supplying options in the second argument.
import { matchPattern } from 'browser-extension-url-match'
const options = {
supportedSchemes: ['http', 'https', 'ftp', 'ftps'],
}
matchPattern('ftps://*/*', options)
.assertValid()
.match('ftps://example.com/foo/bar')
// ⇒ true
The available configuration options are as follows:
If set to false
, the specified path segment is ignored and is always treated as /*
. This corresponds to the behavior when specifying host permissions.
Default: true
.
An array of schemes to allow in the pattern. Available schemes are http
, https
, ws
, wss
, ftp
, ftps
, and file
.
data
and urn
are not currently supported, due to limited implementation and unclear semantics.
Default: ['http', 'https', 'file', 'ftp']
If true
, *
in the scheme will match ws
and wss
as well as http
and https
, which is the default behavior in Firefox.
Default: false
Presets are available to provide defaults based on what Chrome and Firefox support.
import { matchPattern, presets } from 'browser-extension-url-match'
const matcher = matchPattern('*://example.com/', presets.firefox)
matcher.assertValid().match('ws://example.com') // ⇒ true
You can also combine presets with custom options:
const options = {
...presets.firefox,
strict: false,
}
matchPattern('wss://example.com/', options)
.assertValid()
.match('wss://example.com/foo/bar')
// ⇒ true
You can also generate an array of examples matching URL strings from a valid Matcher
object.
matchPattern('https://*.example.com/*').assertValid().examples
// ⇒ [
// 'https://example.com/',
// 'https://example.com/foo',
// 'https://example.com/bar/baz/',
// 'https://www.example.com/',
// 'https://www.example.com/foo',
// 'https://www.example.com/bar/baz/',
// 'https://foo.bar.example.com/',
// 'https://foo.bar.example.com/foo',
// 'https://foo.bar.example.com/bar/baz/',
// ]