Question

I thought it would be neat to parse IRC messages with a regular expression. I got as far as this:

(?::(?<Prefix>[^ ]+) +)?(?<Command>[^ :]+)(?<middle>(?: +[^ :]+)*)(?<coda> +:(?<trailing>.*)?)?

Then I use this with the following .NET code to get the salient elements of the message:

Prefix = matches.Groups["Prefix"].Value;
Command = matches.Groups["Command"].Value;

var parameters = new List<string>();
parameters.AddRange(matches.Groups["middle"].Value
    .Split(new[] { " " }, StringSplitOptions.RemoveEmptyEntries));
parameters.Add(matches.Groups["trailing"].Value);

Parameters = parameters.ToArray();

But I don’t like that I have to split it separately in code. Is there a way that I can obtain an array of matches from the middle group?

Answer 1

You could use the Captures property of a repeated group, although I wouldn’t advice it.

First you would need to change your pattern to:

@"(?::(?<Prefix>[^ ]+) +)?(?<Command>[^ :]+)(?<middle>(?: +[^ :]+))*(?<coda> +:(?<trailing>.*)?)?"

Secondly, you would do:

parameters.AddRange(match.Groups["middle"].Captures.
                    OfType<Capture>().
                    Select(c => c.Value));