Coder Perfect

In a.NET Regex, how do I access named capturing groups?


I’m having trouble finding a good resource that teaches Named Capturing Groups in C#. So far, I’ve got the following code:

string page = Encoding.ASCII.GetString(bytePage);
Regex qariRegex = new Regex("<td><a href=\"(?<link>.*?)\">(?<name>.*?)</a></td>");
MatchCollection mc = qariRegex.Matches(page);
CaptureCollection cc = mc[0].Captures;

However, this always displays the entire line:

<td><a href="/path/to/file">Name of File</a></td> 

I’ve tried a few different “methods” that I’ve discovered on other websites, but they all produce the same effect.

How can I get access to the named capturing groups that my regex specifies?

Asked by UnkwnTech

Solution #1

Use the Match object’s group collection, indexing it with the capturing group name, for example.

foreach (Match m in mc){

Answered by Paolo Tedesco

Solution #2

The named capture group string is specified by supplying it to the indexer of a Match object’s Groups property.

As an example, consider the following:

using System;
using System.Text.RegularExpressions;

class Program
    static void Main()
        String sample = "hello-world-";
        Regex regex = new Regex("-(?<test>[^-]*)-");

        Match match = regex.Match(sample);

        if (match.Success)

Answered by Andrew Hare

Solution #3

Even if there are space characters in between, the following code sample will match the pattern. e.g. :

<td><a href='/path/to/file'>Name of File</a></td>

as well as:

<td> <a      href='/path/to/file' >Name of File</a>  </td>

Whether the input htmlTd string matches the pattern or not, the method returns true or false. If it matches, the link and name are stored in the out params.

/// <summary>
/// Assigns proper values to link and name, if the htmlId matches the pattern
/// </summary>
/// <returns>true if success, false otherwise</returns>
public static bool TryGetHrefDetails(string htmlTd, out string link, out string name)
    link = null;
    name = null;

    string pattern = "<td>\\s*<a\\s*href\\s*=\\s*(?:\"(?<link>[^\"]*)\"|(?<link>\\S+))\\s*>(?<name>.*)\\s*</a>\\s*</td>";

    if (Regex.IsMatch(htmlTd, pattern))
        Regex r = new Regex(pattern,  RegexOptions.IgnoreCase | RegexOptions.Compiled);
        link = r.Match(htmlTd).Result("${link}");
        name = r.Match(htmlTd).Result("${name}");
        return true;
        return false;

This has been tested and shown to be correct.

Answered by SO User

Solution #4

Additionally, if someone has a requirement for group names before running a Regex object search, they can use:

var regex = new Regex(pattern); // initialized somewhere
// ...
var groupNames = regex.GetGroupNames();

Answered by tinamou

Solution #5

This response improves on Rashmi Pandit’s response, which is superior to the others because it appears to fully tackle the problem posed in the question.

The bad news is that it is inefficient and does not consistently employ the IgnoreCase option.

The inefficiency comes from the fact that regex can be costly to generate and execute, and it could have been constructed just once in that answer (by invoking Regex. Behind the scenes, IsMatch was re-constructing the regex). And the Match method might have been run only once and the result saved in a variable, with linkand name calling the Result from that variable.

In addition, the IgnoreCase option was only utilized in the Match section, not in the Regex.IsMatch section.

I also put the Regex definition outside the method so that we only have to construct it once (which I believe is the best approach if we’re saving the assembly with the RegexOptions). Options that have been compiled

private static Regex hrefRegex = new Regex("<td>\\s*<a\\s*href\\s*=\\s*(?:\"(?<link>[^\"]*)\"|(?<link>\\S+))\\s*>(?<name>.*)\\s*</a>\\s*</td>",  RegexOptions.IgnoreCase | RegexOptions.Compiled);

public static bool TryGetHrefDetails(string htmlTd, out string link, out string name)
    var matches = hrefRegex.Match(htmlTd);
    if (matches.Success)
        link = matches.Result("${link}");
        name = matches.Result("${name}");
        return true;
        link = null;
        name = null;
        return false;

Answered by Mariano Desanze

Post is based on