you can find some detailed information about creating a custom extraction rule on the following site http://blogs.msnd.com/joshch as far as parsing out your text goes:

Here is an expample of looking for a tag with a specific attribute. Then grabbing the value from one of the attributes and parsing it to get a specific part of te string. The tag in question will look like the following

<a id="Categories_repCategories_ctl01_lnkCategory" href="Products.aspx page=0&amp;categoryId=BIRDS">Birds</a>

and the code to get the category id out of it looks like this:

  public class ExtractRandomCategory : ExtractionRule
  {
    public override void Extract(object sender, ExtractionEventArgs e)
    {
      List linkTags = new List();
      e.Success = false;
      if (e.Response.IsHtml)
      {
        foreach (HtmlTag tag in e.Response.HtmlDocument.GetFilteredHtmlTags( "a" ))
        {
          if (tag.GetAttributeValueAsString("id").StartsWith("Categories"))
          {
            linkTags.Add(tag);
          }
        }
      }
                                                                          
      Random rand = new Random();
      int linkIndex = rand.Next(linkTags.Count);

      HtmlTag linkTag = linkTags[linkIndex];
      string linkValue = linkTag.GetAttributeValueAsString("href");
      string categoryId;
      string categoryIdParamName = "categoryId=";
      if (linkValue.IndexOf(categoryIdParamName) != -1)
      {
        categoryId = linkValue.Substring((linkValue.IndexOf(categoryIdParamName) + categoryIdParamName.Length));
        e.WebTest.Context.Add(this.ContextParameterName, categoryId);
        e.Success = true;
      }
      else
      {
        e.Success = false;
      }
                                                                                                             
      
      return;

    }

    public override string RuleName
    {
      get { return "Extract Ramdom Category Id"; }
    }
  }