A tiny (399B) utility that converts route patterns into RegExp. Limited alternative to
path-to-regexp
🙇
Withregexparam
,you may turn a pathing string (eg,/users/:id
) into a regular expression.
An object with shape of{ keys, pattern }
is returned, wherepattern
is theRegExp
andkeys
is an array of your parameter name(s) in the order that they appeared.
Unlikepath-to-regexp
,this module does not create akeys
dictionary, nor mutate an existing variable. Also, this only ships a parser, which only accept strings. Similarly, and most importantly,regexparam
onlyhandles basic pathing operators:
- Static (
/foo
,/foo/bar
) - Parameter (
/:title
,/books/:title
,/books/:genre/:title
) - Parameter w/ Suffix (
/movies/:title.mp4
,/movies/:title.(mp4|mov)
) - Optional Parameters (
/:title?
,/books/:title?
,/books/:genre/:title?
) - Wildcards (
*
,/books/*
,/books/:genre/*
) - Optional Wildcard (
/books/*?
)
This module exposes three module definitions:
- CommonJS:
dist/index.js
- ESModule:
dist/index.mjs
- UMD:
dist/index.min.js
$ npm install --save regexparam
import{parse,inject}from'regexparam';
// Example param-assignment
functionexec(path,result){
leti=0,out={};
letmatches=result.pattern.exec(path);
while(i<result.keys.length){
out[result.keys[i]]=matches[++i]||null;
}
returnout;
}
// Parameter, with Optional Parameter
// ---
letfoo=parse('/books/:genre/:title?')
// foo.pattern => /^\/books\/([^\/]+?)(?:\/([^\/]+?))?\/?$/i
// foo.keys => ['genre', 'title']
foo.pattern.test('/books/horror');//=> true
foo.pattern.test('/books/horror/goosebumps');//=> true
exec('/books/horror',foo);
//=> { genre: 'horror', title: null }
exec('/books/horror/goosebumps',foo);
//=> { genre: 'horror', title: 'goosebumps' }
// Parameter, with suffix
// ---
letbar=parse('/movies/:title.(mp4|mov)');
// bar.pattern => /^\/movies\/([^\/]+?)\.(mp4|mov)\/?$/i
// bar.keys => ['title']
bar.pattern.test('/movies/narnia');//=> false
bar.pattern.test('/movies/narnia.mp3');//=> false
bar.pattern.test('/movies/narnia.mp4');//=> true
exec('/movies/narnia.mp4',bar);
//=> { title: 'narnia' }
// Wildcard
// ---
letbaz=parse('users/*');
// baz.pattern => /^\/users\/(.*)\/?$/i
// baz.keys => ['*']
baz.pattern.test('/users');//=> false
baz.pattern.test('/users/lukeed');//=> true
baz.pattern.test('/users/');//=> true
// Optional Wildcard
// ---
letbaz=parse('/users/*?');
// baz.pattern => /^\/users(?:\/(.*))?(?=$|\/)/i
// baz.keys => ['*']
baz.pattern.test('/users');//=> true
baz.pattern.test('/users/lukeed');//=> true
baz.pattern.test('/users/');//=> true
// Injecting
// ---
inject('/users/:id',{
id:'lukeed'
});//=> '/users/lukeed'
inject('/movies/:title.mp4',{
title:'narnia'
});//=> '/movies/narnia.mp4'
inject('/:foo/:bar?/:baz?',{
foo:'aaa'
});//=> '/aaa'
inject('/:foo/:bar?/:baz?',{
foo:'aaa',
baz:'ccc'
});//=> '/aaa/ccc'
inject('/posts/:slug/*',{
slug:'hello',
});//=> '/posts/hello'
inject('/posts/:slug/*',{
slug:'hello',
'*':'x/y/z',
});//=> '/posts/hello/x/y/z'
// Missing non-optional value
// ~> keeps the pattern in output
inject('/hello/:world',{
abc:123
});//=> '/hello/:world'
Important:When matching/testing against a generated RegExp, your pathmustbegin with a leading slash (
"/"
)!
For fine-tuned control, you may pass aRegExp
value directly toregexparam
as its only parameter.
In these situations,regexparam
does notparse nor manipulate your pattern in any way! Because of this,regexparam
has no "insight" on your route, and instead trusts your input fully. In code, this means that the return value'skeys
is always equal tofalse
and thepattern
is identical to your input value.
This also means that you must manage and parse your ownkeys
~!
You may usenamed capture groupsor traverse the matched segments manually the "old-fashioned" way:
Important:Please check your target browsers' and targetNode.js runtimes' support!
// Named capture group
constnamed=regexparam.parse(/^\/posts[/](?<year>[0-9]{4})[/](?<month>[0-9]{2})[/](?<title>[^\/]+)/i);
const{groups}=named.pattern.exec('/posts/2019/05/hello-world');
console.log(groups);
//=> { year: '2019', month: '05', title: 'hello-world' }
// Widely supported / "Old-fashioned"
constnamed=regexparam.parse(/^\/posts[/]([0-9]{4})[/]([0-9]{2})[/]([^\/]+)/i);
const[url,year,month,title]=named.pattern.exec('/posts/2019/05/hello-world');
console.log(year,month,title);
//=> 2019 05 hello-world
Returns:Object
Parse a route pattern into an equivalent RegExp pattern. Also collects the names of pattern's parameters as akeys
array. Aninput
that's already a RegExp is kept as is, andregexparam
makes no additional insights.
Returns a{ keys, pattern }
object, wherepattern
is always aRegExp
instance andkeys
is eitherfalse
or a list of extracted parameter names.
Important:The
keys
willalwaysbefalse
wheninput
is a RegExp and it willalwaysbe an Array wheninput
is a string.
Type:string
orRegExp
Wheninput
is a string, it's treated as a route pattern and an equivalent RegExp is generated.
Note:It does not matter if
input
strings begin with a/
— it will be added if missing.
Wheninput
is a RegExp, it will be usedas is– no modifications will be made.
Type:boolean
Default:false
Should theRegExp
match URLs that are longer than thestr
pattern itself?
By default, the generatedRegExp
will test that the URL begins andends withthe pattern.
Important:When
input
is a RegExp, theloose
argument is ignored!
const{parse}=require('regexparam');
parse('/users').pattern.test('/users/lukeed');//=> false
parse('/users',true).pattern.test('/users/lukeed');//=> true
parse('/users/:name').pattern.test('/users/lukeed/repos');//=> false
parse('/users/:name',true).pattern.test('/users/lukeed/repos');//=> true
Returns:string
Returns a new string by replacing thepattern
segments/parameters with their matching values.
Important:Named segments (eg,
/:name
) thatdo nothave avalues
match will be kept in the output. This is trueexcept foroptional segments (eg,/:name?
) and wildcard segments (eg,/*
).
Type:string
The route pattern that to receive injections.
Type:Record<string, string>
The values to be injected. The keys withinvalues
must match thepattern
's segments in order to be replaced.
Note:To replace a wildcard segment (eg,
/*
), define avalues['*']
key.
As of version1.3.0
,you may useregexparam
with Deno. These options are all valid:
// The official Deno registry:
importregexparamfrom'https://deno.land/x/regexparam/src/index.js';
// Third-party CDNs with ESM support:
importregexparamfrom'https://cdn.skypack.dev/regexparam';
importregexparamfrom'https://esm.sh/regexparam';
Note:All registries support versioned URLs, if desired.
The above examples always resolve to the latest published version.
- trouter- A server-side HTTP router that extends from this module.
- matchit- Similar (650B) library, but relies on String comparison instead of
RegExp
s.
MIT ©Luke Edwards