How to split a nested comma separated list using Regex in C#

Question

I have a string that looks like:

AA,BB,CC(DD,EE(FF))

that I want to split at the root level, leaving the nested list intact. So the above string would be split into:

AA, BB, and CC(DD,EE(FF))

I know how to do this without a regular expression, but I'm having a hard time writing a regex that can identify the outer most commas and ignores the commas in the paratheses.

Obviously post should be getting downvotes due to amount of effort demonstrated in it... Maybe you should show what you've tried and explain why searching for something like https://www.bing.com/search?q=regex+matching+braces (i.e. https://stackoverflow.com/questions/14952113/how-can-i-match-nested-brackets-using-regex) did not help... Than at least post has a chance... — Alexei Levenkov, Jan 23 '18 at 04:26

Kobi · Answer 1 · 2018-01-23T07:05:51.193

In .Net you can use balancing groups to capture nested structures. This is probably more complicated than the code you already have though:

(?>
    (?<Open>\()     # Increase stack depth when matching (
    |
    (?<-Open>\))    # Decrease stack depth when matching )
    |
    [^(),]          # Match any other character
    |
    (?(Open),|(?!)) # Allow commas only while inside parentheses.
)+
(?(Open)(?!))  # Ensure we have no extra open parentheses.

Live example on Regex Storm

A simpler (yet uglier) approach that works in all regex flavors is to limit the levels of nesting you support. For example:

One level of parentheses - just AA and AA(B,B): (?:[^,()]|\([^()]*\))+
Two levels, also matches AA(B,B(C,C))): (?:[^,()]|\((?:[^()]|\([^()]*\))*\))+
Three levels would be: (?:[^,()]|\((?:[^()]|\((?:[^()]|\([^()]*\))*\))*\))+

Basically we're adding another (?:[^()]|\([^()]*\))* for each new level we want to match.

How to split a nested comma separated list using Regex in C#

1 Answers1