public void CleanContentTrusted(AtomContent content)
{
Logger.Info("Cleaning content to be valid xhtml.");
string text = content.Text;
if (content.Type == "html")
{
text = "<div xmlns=\"" + Atom.XhtmlNs.NamespaceName + "\">" + text + "</div>";
}
Tidy tidy = new Tidy();
/* Set the options you want */
tidy.Options.DocType = DocType.Strict;
//tidy.Options.DropFontTags = true;
tidy.Options.LogicalEmphasis = true;
tidy.Options.Xhtml = true;
tidy.Options.XmlOut = true;
tidy.Options.MakeClean = true;
tidy.Options.TidyMark = false;
tidy.Options.QuoteNbsp = false;
tidy.Options.NumEntities = true;
tidy.Options.CharEncoding = CharEncoding.UTF8;
tidy.Options.FixBackslash = true;
tidy.Options.FixComments = true;
TidyMessageCollection tmc = new TidyMessageCollection();
using (MemoryStream input = new MemoryStream())
using (MemoryStream output = new MemoryStream())
{
byte[] bytes = Encoding.UTF8.GetBytes(text);
input.Write(bytes, 0, bytes.Length);
input.Position = 0;
tidy.Parse(input, output, tmc);
text = Encoding.UTF8.GetString(output.ToArray());
if (string.IsNullOrEmpty(text)) throw new FailedToCleanContentException(
string.Format("{0} HTML Tidy Error(s)" + Environment.NewLine, tmc.Errors)
+ string.Join(Environment.NewLine,
tmc.Cast<TidyMessage>()
.Where(m => m.Level == MessageLevel.Error)
.Select(m => m.ToString()).ToArray()));
}
//remove html/body that gets added by tidy
//int start = text.IndexOf("<div xmlns");
//text = text.Substring(start, text.IndexOf("</body>") - start);
XElement div = XElement.Parse(text).Element(Atom.XhtmlNs + "body").Element(Atom.XhtmlNs + "div");
//remove decendent xmlns that gets added by tidy
div.Descendants().Where(d => d.Attribute("xmlns") != null &&
d.Attribute("xmlns").Value == Atom.XhtmlNs.NamespaceName)
.Select(d => d.Attribute("xmlns")).Remove();
//text = text.Replace(" xmlns=\"http://www.w3.org/1999/xhtml\"", string.Empty);
//text = "<div xmlns=\"http://www.w3.org/1999/xhtml\">" + text.Substring("<div>".Length);
//set as xhtml
content.Type = "xhtml";
content.Text = div.ToString(SaveOptions.None);
}
public void FullCreateAtomContentTest()
{
AtomContent atomContent = new AtomContent()
{
Base = new Uri("http://base.com"),
Lang = "EN",
Src = new Uri("http://src.com"),
Text = "This is the text",
Type = "aType"
};
Assert.IsNotNull(atomContent);
Assert.AreEqual(new Uri("http://base.com"), atomContent.Base);
Assert.AreEqual("EN", atomContent.Lang);
Assert.AreEqual(new Uri("http://src.com"), atomContent.Src);
Assert.AreEqual("This is the text", atomContent.Text);
Assert.AreEqual("aType", atomContent.Type);
}
//protected void Validate(AtomContent content)
//{
// XhtmlValidator validator = new XhtmlValidator(content.Text);
// IEnumerable<ValidationRecord> errors = validator.Validate()
// .Where(r => r.Severity == System.Xml.Schema.XmlSeverityType.Error);
// if (errors.Count() > 0)
// {
// throw new Exception("Content is not valid xhtml 1.0 strict:" + Environment.NewLine
// + string.Join(Environment.NewLine, errors.Select(e => e.ToString()).ToArray()));
// }
//}
public void CleanContentFully(AtomContent content)
{
CleanContentTrusted(content);
XElement root = content.Xml.Element(Atom.XhtmlNs + "div");
//remove invalid html elements and attributes
RemoveInvalidNodes(root);
//remove invalid href, src values
foreach (XElement anchor in root.Descendants().Where(d => d.Name.LocalName == "a"))
{
XAttribute href = anchor.Attribute("href");
if (href != null && href.Value.StartsWith("http://") || href.Value.StartsWith("https://")
|| href.Value.StartsWith("ftp://"))
{
Logger.Info("Found valid link: " + anchor.Value);
}
else href.Value = string.Empty;
}
foreach (XElement img in root.Descendants().Where(d => d.Name.LocalName == "img"))
{
XAttribute src = img.Attribute("src");
if (src != null && src.Value.StartsWith("http://") || src.Value.StartsWith("https://"))
{
Logger.Info("Found valid img: " + img.Value);
}
else src.Value = string.Empty;
}
}
public void SimpleCreateAtomContentTest()
{
AtomContent atomContent = new AtomContent();
Assert.IsNotNull(atomContent);
}
