Using Regex in C# to parse a string into an array/list by delimiter with one delimiter exception

Question

My input string:

var s = "{1, >, [4,6,7,8], a, b, [x,y], d, 9}";

I'd like to remove the {} and get an array with each element separated by a comma EXCEPT when the comma is inside [] -- basically anything in brackets will be returned as its own element without the brackets.

Desired output List< String > or String[] whose contents are:

1
>
4,6,7,8
a
b
x,y
d
9

ETA: Here is my UnitText (xunit), which tests each of the patterns suggested by @washington-guedes, with a parameter to trim the input string of whitespaces. The test fails on both cases when the WS is cleaned.

    [Theory]
    [InlineData(@"([^{\s]+(?=(?:,|})))", false)]
    [InlineData(@"([^{\s]+(?=(?:,|})))", true)]
    [InlineData(@"([^{\s[\]]+(?=(?:]|,|})))", false)]
    [InlineData(@"([^{\s[\]]+(?=(?:]|,|})))", true)]
    public void SO(string pattern, bool trimWS)
    {
        //Arrange
        var exp = "{1, >, [4,6,7,8], a, b, [x,y], d, 9}";
        if (trimWS)
            exp = exp.Replace(" ", "");
        Match match = Regex.Match(exp, pattern);
        var list = new List<String>();
        while (match.Success)
        {
            list.Add(match.Value);
            match = match.NextMatch();
        }
        Assert.Equal(8, list.Count);
    }

Can sets of [...] be nested? And are there always spaces between the different major elements? Because right now, it seems like you can just remove the starting '{' and ending '}', split by ", ", and then remove any '['s and ']'s. — Jashaszun
– Jashaszun, Commented Jul 24, 2015 at 21:11
What have you tried? I'm glad to help, but StackExchange isn't a coding service, you're not supposed to say "here's my problem, go solve it." Instead, show us your proposed solution and we can help you refine it. — dmeglio
– dmeglio, Commented Jul 24, 2015 at 21:13

score 1 · Accepted Answer · 2015-07-25 17:18:42Z

1

Try this Regex:

((?<=,\[)[^]]+)|((?<={)[^,}]+)|((?<=,)(?!\[)[^,}]+)

Regex live here.

Explaining:

(                  # start of capturing group
  (?<=,\[)         # starting with ",["
  [^]]+            # matches all till next "]"
)                  # end of capturing group

  |                # OR

(
    (?<={)         # starting with "{"
    [^,}]+         # matches all till next "," or "}"
)

  |                # OR

(
    (?<=,)(?!\[)   # starting with "," and not a "["
    [^,}]+         # matches all till next "," or "}"
)

Hope it helps.

edited Jul 25, 2015 at 17:18

answered Jul 24, 2015 at 21:13

user4227915

Sign up to request clarification or add additional context in comments.

5 Comments

Jashaszun Over a year ago

Regex101 says that this specifically has a non-capturing group, which doesn't seem very useful... shouldn't there be capturing groups in there somewhere?

brandonscript Over a year ago

([^{\s]+(?=(?:,|}))) might work best - it's got the capture groups to make an array.

WhiskerBiscuit Over a year ago

This matches fairly well, however is there a way to ignore whitespace?

WhiskerBiscuit Over a year ago

If you s = s.Replace(" ", ""); before calling the regex, it returns a different set of matches

WhiskerBiscuit Over a year ago

That works great. I was also looking for the version that included {}. I wish I could "read" your expression to understand it. I pasted that expression in regex101. Is the first "part" of the expression that you just added ((?<=,\[)[^]]+)

Collectives™ on Stack Overflow

Using Regex in C# to parse a string into an array/list by delimiter with one delimiter exception

1 Answer 1

5 Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

5 Comments

Your Answer

Sign up or log in

Post as a guest

Related