Unable to build regex to separate path of URL and also query parameters

Question

I built a regex with capture groups to separate path of a URL:
Regex:

(?:\/)(?:([^\/]*))

So the above regex works as follows:
If the URL path is: /some/url/path

The above regex results in:

Match 1: 
    Full match:  /some
    Group 1: some

Match 2:  
    Full match: /url
    Group 1: url

Match 3:  
    Full match: /path
    Group 1: path

This works fine.
But now I also need a regex to parse a URL path with query parameters:
For ex, if the url path is: /some/url/path?name=xyz&age=21&weight=97

The result should be:

Match 1: 
    Full match:  /some
    Group 1: some

Match 2:  
    Full match: /url
    Group 1: url

Match 3:  
    Full match: /path
    Group 1: path

Match 4: 
    Full match: name=xyz
    Group 2: name
    Group 3: xyz

Match 5: 
    Full match: age=21
    Group 2: age
    Group 3: 21

Match 6: 
    Full match: weight=97
    Group 2: weight
    Group 3: 97

Just for information: I am using (regex101) for building regex.

Sounds like you need a proper url parser. Regex isn't powerful enough to parse urls correctly. — Not a real meerkat
– Not a real meerkat, Commented Sep 15, 2018 at 15:15
@Toto I am using nodejs. I am actually contributing to pre-existing code. Which was good at doing the 1st part. — Aditya Singh
– Aditya Singh, Commented Sep 15, 2018 at 15:38

mvp · Accepted Answer · 2018-09-15 17:55:09Z

1

Do not use regex for this, you will get something wrong. Instead, use standard nodejs API: https://nodejs.org/api/url.html

Code example:

const { URL } = require('url');
url = new URL('https://alice:[email protected]:1234/foo?x=1&y=2#bar');
console.log(url)

prints:

URL {
  href: 'https://alice:[email protected]:1234/foo?x=1&y=2#bar',
  origin: 'https://example.org:1234',
  protocol: 'https:',
  username: 'alice',
  password: 'pwd',
  host: 'example.org:1234',
  hostname: 'example.org',
  port: '1234',
  pathname: '/foo',
  search: '?x=1&y=2',
  searchParams: URLSearchParams { 'x' => '1', 'y' => '2' },
  hash: '#bar' }

Good luck trying to do same quality of URL parsing with regexes.

edited Sep 15, 2018 at 17:55

answered Sep 15, 2018 at 15:57

mvp

118k15 gold badges132 silver badges155 bronze badges

Sign up to request clarification or add additional context in comments.

Comments

clabe45 · Accepted Answer · 2018-09-15 15:54:50Z

0

If you can, I would use logic combined with regex, instead of large patterns that don't make much sense. Also, it is impossible to process this kind of url with one regex, if you want the result in multiple matches.

If the path includes a ?, split the string by ?. Process the first substring with your current regex pattern. Then, process the second substring with a different pattern. I think it makes sense to split the param substring by &, and then use regex on each part. Here's a python example

if '?' in url:
    path, params = url.split('?')
    # (process `path` like you are--iteratively search with your current pattern)
    # process `params`
    for part in params.split('&'):
        pattern = '[^=]+=.+'
        # (process `part` with the above pattern)
else:
    # (no query parameters)

With that said, this is very simple. There are url parsers that would do this job much better. ...

edited Sep 15, 2018 at 15:54

answered Sep 15, 2018 at 15:49

clabe45

2,47418 silver badges27 bronze badges

2 Comments

Not a real meerkat Over a year ago

It's not a good idea to split on '?' or '&'. The query string may contain embedded '?'s or escaped '&'s. Granted, my answer suffers from the same problem. A URL parser is really the best match for this (pun intended :D).

clabe45 Over a year ago

@CássioRenan Actually, I believe ? escaped is %3F and & escaped is %26. Nonetheless a url parser is probably better.

Not a real meerkat · Accepted Answer · 2018-09-15 15:56:47Z

0

This is the closest you can get:

(?:\/)(?:([^\/?]*))|[?&]([^&]+)=([^&]+)

But it still has some problems:

It will misbehave if the url passed isn't well-formed
It can't handle escapes in the query string
It doesn't handle the fragment portion of the url.

You are better off using a proper url parser. There are several options available for those.

edited Sep 15, 2018 at 15:56

answered Sep 15, 2018 at 15:49

Not a real meerkat

5,7671 gold badge27 silver badges58 bronze badges

Collectives™ on Stack Overflow

Unable to build regex to separate path of URL and also query parameters

3 Answers 3

Comments

2 Comments

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

3 Answers 3

Comments

2 Comments

Comments

Your Answer

Sign up or log in

Post as a guest

Related