Java Regex : Find out nested patterns and repeated occurrences

Question

I am trying to parse String that has the following patterns:

a2[u]
3[rst]5[g]
3[r2[g]]

I want to extract these strings into following tokens:

2 [u]
3 [rst], 5 [g]
2 [r, 3 [r2[g]] (nested groups)

I am using the following Pattern and Code:

Pattern MY_PATTERN = Pattern.compile("(\\d+)\\[(.+)\\]");
String input = "3[rst]5[g]";
Matcher m = MY_PATTERN.matcher(input);
while(m.find()) {
    System.out.println(m.group(1) + " " + m.group(2));
}

However, it matches to last occurrence of ] instead of the first and that results in an unexpected results. If I change the pattern to (\\d+)\\[(\\w+)\\], it works but fails for 3[r2[g]]. What changes do I need to make so that it doesn't count the whole string as one match?

If you plan to match more nested levels than 1, the regex will become unwieldly. Else, use `"(\\d+)\\[([^\\]\\[]*(?:\\[[^\\]\\[]*][^\\]\\[]*)*)]"`. — Wiktor Stribiżew, Nov 19 '20 at 19:57
@Darshan: I would suggest against using regex for this. Better you use a token parser since you are dealing nested brackets. — anubhava, Nov 20 '20 at 05:13
@WiktorStribiżew It would need more nested levels I am afraid. So, I would go with a token parser as anubhava suggested. — Darshan Mehta, Nov 20 '20 at 08:44

score -1 · Answer 1 · answered Nov 19 '20 at 20:15

-1

Looks like you need to add a quantifier to the .+

As it stands the . will eat the whole string and then only match on the last ]. Add a reluctant quantifier ? to the .+ so make the regex (\\d+)\\[(.+?)\\] and see how far you get...

answered Nov 19 '20 at 20:15

Michael Wiles

20,902
18
71
101

1

No way, lazy dot will match `3[r2[g]` in `3[r2[g]]`. – Wiktor Stribiżew Nov 19 '20 at 20:59
okay - so regex is not the answer then – Michael Wiles Nov 20 '20 at 09:21

Java Regex : Find out nested patterns and repeated occurrences

1 Answers1