Make multi line regex non greedy (=lazy)

Question

I have a snippet of code like this:

{
  a: {
    b: 1,
    c: 2,
  },

  d: {
    e: 4,
    f: 5,
  },
}

I need a regex which matches the two objects a and d seperately.

First match:

a: {
  b: 1,
  c: 2,
},

Second match:

d: {
  e: 4,
  f: 5,
},

I tried this:

  .+: \{
([\s\S\r]+)
  \},

But it greedily matches the two objects as one match.

How do I make the expression lazy?

I tried to place the lazy operator ? at several places. It didn't work.

I use the text search in Visual Studio Code.

Parse the whole thing as [JSON5](https://json5.org/), then extract whatever you need from that. — InSync, Jul 20 '23 at 14:46
Are the leading spaces actually part of the regular expression? — Peter Mortensen, Jul 23 '23 at 09:36

score -1 · Answer 1 · edited Jul 23 '23 at 09:21

-1

Thank you, marchant felts. Your answer led me to the solution:

  .+: \{
([\s\S\r]+?)
  \},

I added the ? in the second line after the +. It works in Visual Studio Code.

edited Jul 23 '23 at 09:21

Peter Mortensen

30,738
21
105
131

answered Jul 20 '23 at 12:10

Jochen Kunze

583
1
5
17

score -2 · Answer 2 · answered Jul 20 '23 at 11:40

In the regex you provided, there seem to be some mismatched components compared to your code sample. In your code sample, there are no "catalog_ids" properties. Let's modify the regex to conform to your code sample.

In order to make your regex non-greedy (or lazy), you need to employ the ? character in the right place. To capture the JavaScript objects you mentioned, you should position the ? character just behind the + inside your group, like (.+?). This means "match any character one or more times, but as few as possible".

Here is a regex that should work:

(\w+):\s+\{[\s\S]*?\},

Explanation:

(\w+) - matches any word character (equal to [a-zA-Z0-9_])
:\s+\{ - matches : then any whitespace characters \s+, then {
[\s\S]*? - matches any character (including line breaks), but as few as possible due to the ?
\}, - matches },

Please bear in mind that regex isn't necessarily the best tool to parse nested structures like JSON. Depending on the language you're working in, you may want to consider utilizing a JSON parser or similar tool instead.

As you're employing VS Code, this regex should work fine if you ensure that the "Use Regular Expressions" and "Match Case" options are checked in the search panel. Also remember that regex in the search bar of VS Code doesn't necessitate surrounding / delimiters.

This answer is likely ChatGPT with some synonyms replaced (like all this user's other answers) — DavidW, Jul 23 '23 at 05:58
[Another answer](https://stackoverflow.com/questions/76744618/typescript-stage-3-fields-decorators/76744632#76744632) had a slip-up: *"At the epoch of my knowledge cutoff in September 2021"*. — Peter Mortensen, Jul 23 '23 at 09:53

Make multi line regex non greedy (=lazy)

2 Answers2