C# Convert Relative to Absolute Links in HTML String

Question

I'm mirroring some internal websites for backup purposes. As of right now I basically use this c# code:

System.Net.WebClient client = new System.Net.WebClient();
byte[] dl = client.DownloadData(url);

This just basically downloads the html and into a byte array. This is what I want. The problem however is that the links within the html are most of the time relative, not absolute.

I basically want to append whatever the full http://domain.is before the relative link as to convert it to an absolute link that will redirect to the original content. I'm basically just concerned with href= and src=. Is there a regex expression that will cover some of the basic cases?

Edit [My Attempt]:

public static string RelativeToAbsoluteURLS(string text, string absoluteUrl)
{
    if (String.IsNullOrEmpty(text))
    {
        return text;
    }

    String value = Regex.Replace(
        text, 
        "<(.*?)(src|href)=\"(?!http)(.*?)\"(.*?)>", 
        "<$1$2=\"" + absoluteUrl + "$3\"$4>", 
        RegexOptions.IgnoreCase | RegexOptions.Multiline);

    return value.Replace(absoluteUrl + "/", absoluteUrl);
}

Answer 1

The most robust solution would be to use the HTMLAgilityPack as others have suggested. However a reasonable solution using regular expressions is possible using the Replace overload that takes a MatchEvaluator delegate, as follows:

var baseUri = new Uri("http://test.com");
var pattern = @"(?<name>src|href)=""(?<value>/[^""]*)""";
var matchEvaluator = new MatchEvaluator(
    match =>
    {
        var value = match.Groups["value"].Value;
        Uri uri;

        if (Uri.TryCreate(baseUri, value, out uri))
        {
            var name = match.Groups["name"].Value;
            return string.Format("{0}=\"{1}\"", name, uri.AbsoluteUri);
        }

        return null;
    });
var adjustedHtml = Regex.Replace(originalHtml, pattern, matchEvaluator);

The above sample searches for attributes named src and href that contain double quoted values starting with a forward slash. For each match, the static Uri.TryCreate method is used to determine if the value is a valid relative uri.

Note that this solution doesn't handle single quoted attribute values and certainly doesn't work on poorly formed HTML with unquoted values.