using System;
using System.Collections;
using HtmlAgilityPack;

public sealed class SafeHtml {

	private SafeHtml() {}

	private static readonly string[] allowedElements = 
		new string[] {
						 // #DOCUMENT and #TEXT are HtmlAgilityPack names
						 "!DOCTYPE", "#DOCUMENT", "#TEXT",	
						 "A", "ACRONYM", "ADDRESS", "AREA", "B", "BASE", 
						 "BASEFONT", "BDO", "BIG", "BLOCKQUOTE", "BODY", "BR", "CAPTION",
						 "CENTER", "CITE", "CODE", "COL", "COLGROUP", "DD", "DEL", "DFN",
						 "DIR", "DIV", "DL", "EM", "FONT", "HEAD", "H1", "H2", "H3", "H4",
						 "H5", "H6", "HR", "HTML", "I", "IMG", "INS", "KBD", "LABEL",
						 "LI", "LISTING", "MAP", "MARQUEE", "MENU", "NOBR", "NOSCRIPT",
						 "OL", "P", "PRE", "Q", "RT", "RUBY", "S", "SAMP", "SMALL", "SPAN",
						 "STRIKE", "STRONG", "SUB", "SUP", "TABLE", "TBODY", "TD", "TFOOT",
						 "TH", "THEAD", "TITLE", "TR", "TT", "U", "UL", "VAR", "WBR", "XMP"
					 };

	private static readonly string[] allowedAttributes = 
		new string[] {
						 "ABBR", "ADDITIVE", "ALIGN", "ALLOWTRANSPARENCY", "ALT", "ATOMICSELECTION",
						 "AXIS", "BACKGROUND", "BGCOLOR", "BGPROPERTIES", "BORDER", "BORDERCOLOR",
						 "CAPTION", "CELLPADDING", "CELLSPACING", "CITE", "CLEAR", "COLOR", 
						 "COLS", "COLSPAN", "COMPACT", "COORDS", "DIR", "DISABLED", "DYNSRC", 
						 "FACE", "FGCOLOR", "FONT", "HEADERS", "HEIGHT", "HREF", "HREFLANG",
						 "HSPACE", "ID", "LENGTH", "LINK", "LOWSRC", "MEDIA", "NAME", "NOHREF",
						 "NOSHADE", "NOWRAP", "REL", "REV", "ROWS", "ROWSPAN", "SCROLL", 
						 "SHAPE", "SIZE", "SPAN", "SRC", "START", "STYLE", "TARGET", "TEXT",
						 "TITLE", "TYPE", "UNSELECTABLE", "URN", "VALIGN", "VALUE", "VSPACE",
						 "WIDTH", "WRAP"
					 };

	public static string CleanHtml(string html) {
		// Load data
		HtmlDocument doc = new HtmlDocument();
		doc.LoadHtml(html);
			
		// Start processing
		processNode(doc.DocumentNode);
		return doc.DocumentNode.ChildNodes[0].WriteTo();
	}


	private static void processNode(HtmlNode node) {
		// Check to see if element is not allowed 
		if (Array.BinarySearch(allowedElements, node.Name.ToUpper()) < 0) {
			// Rename node to something simple, remove all children and attributes
			node.Name = "PRE";
			node.RemoveAll();
			return;
		}

		// Check all attributes, adding bad ones to removeList
		ArrayList removeList = new ArrayList();
		foreach(HtmlAttribute att in node.Attributes) {
			// If attribute is not allowed, add for removal
			if (Array.BinarySearch(allowedAttributes, att.Name.ToUpper()) < 0) {
				removeList.Add(att);
			} else {
				// Check for style, to disable behaviors
				if (att.Name.ToUpper() == "STYLE") {
					cleanStyle(att);
				}
			}
		}

		// Remove attributes
		foreach(HtmlAttribute att in removeList) {
			node.Attributes.Remove(att);
		}

		// Process child nodes
		foreach(HtmlNode child in node.ChildNodes) {
			processNode(child);
		}
	}


	private static void cleanStyle(HtmlAttribute styleAttribute) {
		// HACK: Cheap and quick way to remove positioning and behaviors from a style 
		string val = styleAttribute.Value.ToLower();
		val = val.Replace("behavior", "");
		val = val.Replace("position", "");
		val = val.Replace("top", "");
		val = val.Replace("left", "");
		styleAttribute.Value = val;
	}
}