Question

I sadly have a requirement to generate some messy XML.

The main document must contain an embedded XML document. However, the embedded document occurs within a CDATA section. The final result should look something like this:

<?xml version="1.0"?>
<foo>
  <xml>
    <![CDATA[
      <?xml version="1.0" encoding="UTF-8"?>
      <bar>
      </bar>
    ]]>
  </xml>
</foo>

I am running into two problems:

First, anything that is output within the CDATA section is rendering as escaped (e.g., the greater than sign > becomes >)

Is there a way to disable the escaping within the CDATA section?

Second, I am unable to create the XML declaration. I receive the following exception when attempting to include the embedded XML document:

def serializeEmbedded(): Seq[Node] = {
  <?xml version="1.0"?>
  <bar>
  </bar>
}

Exception in thread "main" java.lang.IllegalArgumentException: xml is reserved
    at scala.xml.ProcInstr.<init>(ProcInstr.scala:25)

This is my first foray into Scala’s native XML processing.

Thank you,

Saish

Answer 1

The XML declaration is only really relevant for serialization, and you can’t specify it using Scala’s XML literal syntax (as you’ve discovered).

I’d suggest defining a helper function like this:

import scala.xml._

def toCData(doc: Elem) = {
  val w = new java.io.StringWriter
  XML.write(w, doc, "UTF-8", true, null)
  PCData(w.toString)
}

Now you can write, for example:

scala> val doc = <outer>{ toCData(<inner/>) }</outer>
doc: scala.xml.Elem = 
<outer><![CDATA[<?xml version='1.0' encoding='UTF-8'?>
<inner></inner>]]></outer>

It’s not as elegant as it could be with a little more language support, but it works.