Regex, I need to improve a method to get src and alt from an image

There is no problem to get the src or alt separately,but how can get both at the same time every one with a group name.

We have to bear in mind that alt can be to the left or right of src.

I am in a hurry, so I found a quick solution, creating 3 groupnames for the src, and and for alt. I know we can do it a lot better.

private void GetFirstImage(string newHtml, out string imgstring, out string imgalt)
{
    imgalt = "";
    imgstring = "";

    string pattern = "(?<=<img(?<name1>\\s+[^>]*?)src=(?<q>['\"]))(?<url>.+?)(?=\\k<q>)(?<name2>.+?)\\s*\\>";

    try
    {
        //si hay imagen
        if (Regex.IsMatch(newHtml, pattern))
        {
            Regex r = new Regex(pattern, RegexOptions.IgnoreCase | RegexOptions.Compiled);

            imgstring = r.Match(newHtml).Result("${url}");
            string tempalt = "", tempalt2;
            tempalt = r.Match(newHtml).Result("${name1}");
            tempalt2 = r.Match(newHtml).Result("${name2}");

            //ya tenemos la ruta de la imagen y de lo que aparece a izq y a derecha dentro de <img>

            try
            {
  开发者_运维技巧              pattern = "alt=(?<q>['\"])(?<alt>.+?)(?=\\k<q>)";

                //si hay algo que no sea vacío a la izquierda de la ruta
                if(!String.IsNullOrEmpty(tempalt.Trim()))
                {
                    r = new Regex(pattern, RegexOptions.IgnoreCase | RegexOptions.Compiled);

                    //si cumple con el pattern para buscar el alt
                    if (Regex.IsMatch(tempalt, pattern))
                    {

                        imgalt = r.Match(tempalt).Result("${alt}");

                    }
                }
                //si no se encontró el alt y hay algo a la derecha
                if(String.IsNullOrEmpty(imgalt) && !String.IsNullOrEmpty(tempalt2.Trim()))
                {

                    r = new Regex(pattern, RegexOptions.IgnoreCase | RegexOptions.Compiled);

                    //si se cumple el patrón del alt
                    if (Regex.IsMatch(tempalt2, pattern))
                    {

                        imgalt = r.Match(tempalt2).Result("${alt}");

                    }

                }

            }
            catch{ }

        }

    }
    catch{}

}

Simple... don't use Regex. Use a DOM parser - so XmlDocument for xhtml or the HTML Agility Pack for (non-x)html.

Then just query root.SelectNodes("//img") and look at the "src" and "alt" attributes on each element (i.e. node.Attributes["src"].Value, etc)

Regex is NOT a good tool for parsing html (since it is not a regular language).