Getting multiple matches in a hierarchical string using regex

Question

Here is my string:

((A_1:2,B:3):2.1,C_3:1.2,(D:3,(E:4.3,F:2):3.1,):G:1.7);

I need to be able to capture any and all sets of parentheses pairs e.g.

(A_1:2,B:3)

and

(D:3,(E:4.3,F:2):3.1,)

This:

\([^ ]+\)

will capture the entire string, but I can't find a combination of group systems that will capture at least 10 levels of nested parentheses pairs.

I hope to be able to put the captured groups into a data structure so I can parse it more easily. But first, I need to capture the pairs.

you want to capture all (...) inside () ? and skip rest yes? — Maciej Kozieja, Feb 11 '17 at 12:52
Possible duplicate of [Can regular expressions be used to match nested patterns?](http://stackoverflow.com/questions/133601/can-regular-expressions-be-used-to-match-nested-patterns) — Zero Piraeus, Feb 11 '17 at 12:56
Well, everything is in at least one (), so there is nothing to skip. — Glubbdrubb, Feb 11 '17 at 12:58
You may get what you need with PyPi regex module. Not with the standard re module. — Wiktor Stribiżew, Feb 11 '17 at 13:13
I edited it. I don't need an that many levels of nested pairs. — Glubbdrubb, Feb 11 '17 at 13:17
What is target output ? array with all matches? Is this ok? `["(A_1:2,B:3)","(D:3,(E:4.3,F:2):3.1,)"]` — Maciej Kozieja, Feb 11 '17 at 14:54

score 1 · Accepted Answer · answered Feb 11 '17 at 15:24

1

I think you don't need regex for this event more it's harder with.
Here's what i came up with:
Asuming your target datatype is array of stings

def split(data):
    temp = ""
    data = data[1:-2]
    array = []
    closed = opened = 0
    for letter in data:
        if letter == '(':
            opened += 1
        elif letter == ')':
            closed += 1

        if opened != 0:
            temp += letter
            if opened == closed:
                array.append(temp)
                temp = ""
                opened = 0
                closed = 0
    return array

print(split("((A_1:2,B:3):2.1,C_3:1.2,(D:3,(E:4.3,F:2):3.1,):G:1.7);"))

answered Feb 11 '17 at 15:24

Maciej Kozieja

1,812
1
13
32

I can work with an array of strings. This results in: ['(A_1:2,B:3)', '(D:3,(E:4.3,F:2):3.1,)'] (D:3,(E:4.3,F:2):3.1,) still needs to be split off. I assume if I want to go another level, I just run the function again, but on the second element in the array? – Glubbdrubb Feb 11 '17 at 16:13
So is it fine ? I can change it but you have to specify output. – Maciej Kozieja Feb 11 '17 at 16:14
@Glubbdrubb yes it would work but then you have to remove this `data = data[1:-2]` and use is at firs data because for this to work input must be `(A_1:2,B:3):2.1,C_3:1.2,(D:3,(E:4.3,F:2):3.1,):G:1.7` so without first brackets thats done by `data = data[1:-2]` if you want to use this function recursivly then you have to remove this fragment – Maciej Kozieja Feb 11 '17 at 16:25
That's great! :D Would be nice if you accept answer :) – Maciej Kozieja Feb 11 '17 at 16:32

score 0 · Answer 2 · answered Feb 12 '17 at 13:39

0

You cannot do that with regexp matching. The reason is because the language of all matching parenthesized strings is not regular. Look at some bibliography of pattern matching and finite automata, and you'll find the mathematical reasoning behind this.

answered Feb 12 '17 at 13:39

Luis Colorado

10,974
1
16
31

Getting multiple matches in a hierarchical string using regex

2 Answers2