MailMergeLib - A Mail Client Library for .NET

using System.Collections.Generic;
using System.IO;
using System.Text.RegularExpressions;
using HtmlAgilityPack;

namespace MailMergeLib
{
	/// <summary>
	/// Convert HTML to plain text using the HtmlAgilityPack library.
	/// Processing is not too fast, but as it must be done only once
	/// per mail merge job and email texts are usually small, it's okay.
	/// </summary>
	/// <remarks>
	/// Should eventually be improved ;-)
	/// See also http://daringfireball.net/projects/markdown/">Inspired by Markdown
	/// </remarks>
	public class ParsingHtmlConverter : IHtmlConverter
	{
		private const string CrLf = "\r\n";
		private const string CrLfCrLf = "\r\n\r\n";

		#region IHtmlConverter Members

		/// <summary>
		/// Convert a text file with HTML content to plain text.
		/// </summary>
		/// <param name="html">The HTML string to convert.</param>
		/// <returns>The plain text representation of the HTML content.</returns>
		public string ToPlainText(string html)
		{
			var doc = new HtmlDocument();
			doc.LoadHtml(html);
			return ToPlainText(doc);
		}

		#endregion

		private string ToPlainText(HtmlDocument doc)
		{
			var sw = new StringWriter();
			ConvertToText(doc.DocumentNode, sw);
			sw.Flush();
			return sw.ToString();
		}

		private void ConvertContentToText(HtmlNode node, TextWriter outText)
		{
			foreach (HtmlNode subnode in node.ChildNodes)
			{
				ConvertToText(subnode, outText);
			}
		}

		private void ConvertToText(HtmlNode node, TextWriter outText)
		{
			string html;
			switch (node.NodeType)
			{
				case HtmlNodeType.Comment:
					// don't output comments
					break;


				case HtmlNodeType.Document:
					ConvertContentToText(node, outText);
					break;


				case HtmlNodeType.Text:
					string parentName = node.ParentNode.Name;

					// script, style and title text is ignored
					if ((parentName == "script") || (parentName == "style") || parentName == "head" || parentName == "title")
						break;

					html = ((HtmlTextNode) node).Text;

					if (parentName != "pre")
					{
						// get text with all characters remodes which are not visible in html
						html = html.Replace("\t", string.Empty).Replace("\r", string.Empty).Replace("\n", string.Empty);
						var regEx =
							new Regex(@"\s+", RegexOptions.Compiled);
						html = regEx.Replace(html, " ");
					}

					// gracefully handle overlapping closing elements
					if (HtmlNode.IsOverlappedClosingElement(html))
						break;

					html = HtmlEntity.DeEntitize(html);
					if (html.Length > 0)
					{
						outText.Write(html);
					}
					break;


				case HtmlNodeType.Element:
					var toWrite = new List<string>();

					switch (node.Name)
					{
						case "p":
							outText.Write(CrLfCrLf);
							break;
						case "br":
						case "td":
						case "ul":
						case "ol":
							outText.Write(CrLf);
							break;
						case "img":
							// images
							toWrite.Clear();
							if (node.Attributes["src"] != null && node.Attributes["src"].Value.Trim() != string.Empty)
								toWrite.Add("[" + HtmlEntity.DeEntitize(node.Attributes["src"].Value + "]"));
							if (node.Attributes["alt"] != null && node.Attributes["alt"].Value.Trim() != string.Empty)
								toWrite.Add("[" + HtmlEntity.DeEntitize(node.Attributes["alt"].Value + "]"));
							if (node.Attributes["title"] != null && node.Attributes["title"].Value.Trim() != string.Empty)
								toWrite.Add("(\"" + HtmlEntity.DeEntitize(node.Attributes["title"].Value + "\")"));
							outText.Write("[" + string.Join(" ", toWrite.ToArray()) + "] ");
							break;
						case "a":
							// links
							toWrite.Clear();
							if (node.Attributes["href"] != null && node.Attributes["href"].Value.Trim() != string.Empty)
								toWrite.Add("[" + HtmlEntity.DeEntitize(node.Attributes["href"].Value + "]"));
							if (node.Attributes["title"] != null && node.Attributes["title"].Value.Trim() != string.Empty)
								toWrite.Add("(\"" + HtmlEntity.DeEntitize(node.Attributes["title"].Value + "\")"));
							outText.Write(string.Join(" ", toWrite.ToArray()) + " ");
							break;
						case "hr":
							outText.Write("{0}----------{0}", CrLf);
							break;
						case "b":
						case "strong":
							node.InnerHtml = "**" + node.InnerHtml + "**";
							break;
						case "i":
						case "u":
						case "em":
							node.InnerHtml = "_" + node.InnerHtml + "_";
							break;
						case "li":
							node.InnerHtml = "* " + node.InnerHtml + "<br />";
							break;
						case "h1":
						case "h2":
						case "h3":
						case "h4":
						case "h5":
						case "h6":
							// headlines
							outText.Write(CrLfCrLf);
							node.InnerHtml = "#######".Substring(0, 7 - int.Parse(node.Name.Substring(1))) + " " + node.InnerHtml + "<br />";
							break;
					}

					if (node.HasChildNodes)
					{
						ConvertContentToText(node, outText);
					}
					break;
			}
		}
	}
}