RegExp selection in URL

Question

I'm trying to replace tokens in a file. There can be multiple tokens on a line and the delimeters are &&.

Example:

{
  "host": "&&main_service&&/&&auth_endpoint&&"
}

If you use the regex:

const delimiter = '&&';
const delimeterRegex = new RegExp(`${delimiter}.*${delimiter}`);

...the problem is that that doesn't match individually; it can match that whole string (so I get ["&&main_service&&/&&auth_endpoint&&"] as a result, rather than getting ["&&main_service&&", "&&auth_endpoint&&"])

How do I get the two results separately, rather than together?

EDIT: Code I use to do the replace:

const findUnreplacedTokens = () => {
  console.log('Scanning for unreplaced tokens...');
  const errors = [];
  lines.forEach((ln, i) => {
    const tokens = delimeterRegex.exec(ln);
    if (tokens) {
      errors.push({
        tokens,
        line: i+1,
      });
    }
  });

  if (errors.length) {
    handleErrors(errors);
  } else {
    console.log('No rogue tokens found');
    console.log(logSpacing);
  }
};

Possible duplicate of [How to make Regular expression into non-greedy?](https://stackoverflow.com/questions/2824302/how-to-make-regular-expression-into-non-greedy) — Federico klez Culloca, Feb 28 '18 at 16:32
@FedericoklezCulloca I'm afraid not, if it's non-greedy like that, then it only removes the first match, even if I add a 'g' flag to the delimeterRegex. :( — user1775718, Feb 28 '18 at 16:46
IF you go to https://regex101.com/ and have the test string as `&&main_service&&/&&auth_endpoint&&` and the regex as `&&.*?&&*` (the forward slashes and the 'g' flag are added automatically) then that matches. I think it's the end star that is missing. — user1775718, Feb 28 '18 at 17:08
@revo No. Not yet. I'll accept as soon as it is. ;) I've edited to give the execution code, as you asked. — user1775718, Feb 28 '18 at 17:15

zelda11 · Answer 1 · 2018-02-28T16:46:19.353

0

With [^&] and g(global)

const delimeterRegex = new RegExp(`${delimiter}[^&]*${delimiter}`, 'g');

edited Feb 28 '18 at 16:46

answered Feb 28 '18 at 16:36

zelda11

425
2
7

That would mean that the regex would have to match at the start of the string and finish at the end, and that might not be the case. – user1775718 Feb 28 '18 at 16:45
That seems to match the delimeters in the middle. Try it at https://regex101.com/ – user1775718 Feb 28 '18 at 17:08

score 0 · Accepted Answer · answered Feb 28 '18 at 16:53

0

const regex = /&&(.*?)&&/g;
const str = `&&main_service&&/&&auth_endpoint&&`;
let m;

while ((m = regex.exec(str)) !== null) {
    // This is necessary to avoid infinite loops with zero-width matches
    if (m.index === regex.lastIndex) {
        regex.lastIndex++;
    }
    
    // The result can be accessed through the `m`-variable.
    m.forEach(function(match, groupIndex) {
      if(match.indexOf("&&") !== -1) return;
        console.log(`Found match, group ${groupIndex}: ${match}`);
    });
}

answered Feb 28 '18 at 16:53

Ryan

295
2
8

This is correct. The problem I was having came in two parts: (1) The `?` to make it non-greedy (must go inside capture group); (2) MDN says that exec returns an array yet, that's not the whole story - exec acts more like an generator. You need to run the exec repeatedly to get matches. (https://stackoverflow.com/a/9214955/1775718 <-- helped too) – user1775718 Mar 01 '18 at 14:08

RegExp selection in URL

2 Answers2