Regex divide string by commas ignoring function syntax

Question

I need a regex that substitutes a string by looking at their commas.
For example the string:

str1 = "a,b,12,func(a,b),8,bob,func(1,2))"

should be transformed as following:

str1_transformed = "a;b;12;func(a,b);8;bob;func(1,2))"

I cannot substitute every "," with a ";" because it will look like:

str1_wrong = "a;b;12;func(a;b);8;bob;func(1;2))"

Which programming language are you coding it? I've guessed it is C#, if I'm wrong. please retag it. — Dragonthoughts, May 20 '18 at 19:20
@Dragonthoughts I'm using scala, but regex should work indifferently by the program language — Davide, May 20 '18 at 19:22
I agree that there should be no difference, but the actuality is that different languages and libraries implement regex differently. — Dragonthoughts, May 20 '18 at 19:24
You may try `,(?![^()]*\))` to select commas that are outside of brackets. See demo here https://regex101.com/r/8X8TWH/1 — revo, May 20 '18 at 19:27
@revo Thanks. This work great. Now I try to understand it. Thank you so much — Davide, May 20 '18 at 19:29

score 3 · Accepted Answer · answered May 20 '18 at 19:32

3

If you know that you won't have unbalanced or escaped brackets below regex works well:

,(?![^()]*\))

Breakdown:

, Match a comma
(?! Start of negative lookahead
- [^()]*\) That means, recent matched comma shouldn't follow a closing bracket without matching opening bracket
) End of lookahead

C# code:

Regex regex = new Regex(@",(?![^()]*\))");
string result = regex.Replace(@"a,b,12,func(a,b),8,bob,func(1,2))", @";");

answered May 20 '18 at 19:32

revo

Is the `)` inside `[^()]` actually needed? – Andrey Tyukin May 20 '18 at 19:37
1

No it doesn't change result but makes engine to fail fast when more than one `)` character are following each other. – revo May 20 '18 at 19:43

1 Answers1