URL decoding confusion

Question

I\'ve got a DB that refers to the following url:

http://en.wikipedia.org/wiki/Herbert_Gr%F6nemeyer

However, it seems that this is a bad URLEncoding, causing

Answer 1

%C3%B6 is proper UTF-8 encoding for ö (o-umlaut). I would assume that %F6 is byte-for-byte copy of byte value for some local encoding of same character (e.g. from code page 1252).

Answer 2

Here's some quick'n'dirty code I cobbled together to make sense of this. Thanks to Josip for pointing me in the right direction:

    private string UrlDecode(string input)
    {
        string unescaped = null;
        try
        {
            unescaped = Uri.UnescapeDataString(input);
        }
        catch
        {
            unescaped = input;
            for (; ; )
            {
                var match = Regex.Match(unescaped, @"\%[A-F0-9]{2}");
                if (!match.Success)
                    break;
                byte b;
                try
                {
                    b = byte.Parse(match.Value.Substring(1), NumberStyles.HexNumber);
                }
                catch
                {
                    return HttpUtility.UrlDecode(input);
                }
                var replacement = Encoding.GetEncoding(1252).GetString(new[] { b });
                unescaped = unescaped.Substring(0, match.Index) + replacement + unescaped.Substring(match.Index + match.Length);
            }
        }
        return unescaped;
    }