Converting XDocument to JSON with XHTML node using JSON.NET - json

Question

Is there a simpler way to do this? I am sincerely concerned that this LinqPad doodle might be too verbose:
var xml = #"
<root attr1=""attribute one"" attr2=""attribute two"">
<title>this is the root</title>
<xhtml>
<div id=""wrapper"">
<p style=""first""><em>hello</em> world!</p>
</div>
</xhtml>
</root>
";
var xDoc = XDocument.Parse(xml);
Func<XNode, string> getXhtml = node =>
{
var s = string.Empty;
StringWriter wr = null;
try
{
wr = new StringWriter();
using(var jsonWriter = new JsonTextWriter(wr))
{
jsonWriter.StringEscapeHandling = StringEscapeHandling.EscapeHtml;
new JsonSerializer().Serialize(jsonWriter, node.ToString());
}
s = wr.ToString();
}
finally
{
if(wr != null) wr.Dispose();
}
return s;
};
var escaped_xhtml = getXhtml(xDoc.Root.Element("xhtml").Element("div"));
escaped_xhtml.Dump("escaped xhtml");
var placeholder = "#rx-xhtml";
xDoc.Root.Element("xhtml").Value = placeholder;
var json = JsonConvert.SerializeXNode(xDoc.Root, Newtonsoft.Json.Formatting.Indented);
var json_final =json
.Replace(string.Format(#"""{0}""",placeholder), escaped_xhtml);
json_final.Dump("json with escaped xhtml");
var jDoc = JObject.Parse(json_final);
escaped_xhtml = (string)jDoc["root"]["xhtml"];
escaped_xhtml.Dump("decoded xhtml");
It may be reasonable to assume that JSON.NET already has, say, the equivalent of my getXhtml() routine. JSON.NET has so many great features I am concerned that I might be missing out.

Answer 1

Is this what you want? It produces the same result for json_final:
// Parse the XML
var xDoc = XDocument.Parse(xml);
// Extract the "div" element.
var div = xDoc.Root.Element("xhtml").Element("div");
div.Remove();
// Convert the remainder to a JObject.
var jDoc = JObject.FromObject(xDoc.Root, JsonSerializer.CreateDefault(new JsonSerializerSettings { Converters = new[] { new XmlNodeConverter() } }));
// Store the Div element as a string literal.
jDoc["root"]["xhtml"] = div.ToString();
// Stringify the JSON using StringEscapeHandling = StringEscapeHandling.EscapeHtml
var json_final = JsonConvert.SerializeObject(jDoc, new JsonSerializerSettings { Formatting = Formatting.Indented, StringEscapeHandling = StringEscapeHandling.EscapeHtml });
It eliminates the string replacement and the getXhtml delegate. Note that StringEscapeHandling applies during the final conversion to string, not to intermediate conversions to Joken (where string literals are not escaped).
Even simpler still, replace the <Div> element with a corresponding XML string literal before converting to JSON:
// Parse the XML
var xDoc = XDocument.Parse(xml);
// Replace the DIV element with its XML string literal value
var div = xDoc.Root.Element("xhtml").Element("div");
div.Remove();
xDoc.Root.Element("xhtml").Value = div.ToString();
// Convert to JSON using StringEscapeHandling = StringEscapeHandling.EscapeHtml
var json_final = JsonConvert.SerializeObject(xDoc, new JsonSerializerSettings { Converters = new[] { new XmlNodeConverter() }, Formatting = Formatting.Indented, StringEscapeHandling = StringEscapeHandling.EscapeHtml });