Maintain empty/non-elements in stringAsXml output

The stringAsXml implementation currently converts both `<foo/>` and
`<foo></foo>` XML strings to `<foo/>`. Although stringAsXml does not
guarantee the resulting XML will be always be the same, we do try to
keep keep the result as close to the original where reasonable.

By casting the existing XMLReader/Writer classes to XMLStreamReader2 and
XMLStreamWriter2, we can used additional APIs to detect empty elements
and write them as empty, with all other elements written with full
endings.

DAFFODIL-3074

Author: stevedlawrence

daffodil-core/src/main/scala/org/apache/daffodil/runtime1/infoset/XMLTextInfosetInputter.scala

@@ -24,8 +24,6 @@ import javax.xml.XMLConstants
 import javax.xml.stream.XMLInputFactory
 import javax.xml.stream.XMLStreamConstants.*
 import javax.xml.stream.XMLStreamException
-import javax.xml.stream.XMLStreamReader
-import javax.xml.stream.XMLStreamWriter
 
 import org.apache.daffodil.api
 import org.apache.daffodil.api.Daffodil.InfosetInputterEventType
@@ -34,7 +32,9 @@ import org.apache.daffodil.lib.exceptions.Assert
 import org.apache.daffodil.lib.xml.XMLUtils
 import org.apache.daffodil.runtime1.dpath.NodeInfo
 
-import com.ctc.wstx.cfg.ErrorConsts;
+import com.ctc.wstx.cfg.ErrorConsts
+import org.codehaus.stax2.XMLStreamReader2
+import org.codehaus.stax2.XMLStreamWriter2
 
 object XMLTextInfoset {
   lazy val xmlInputFactory = {
@@ -140,10 +140,14 @@ object XMLTextInfoset {
    *
    * Both a lone CR and CRLF are converted to LF.
    */
-  def writeXMLStreamEvent(xsr: XMLStreamReader, xsw: XMLStreamWriter): Unit = {
+  def writeXMLStreamEvent(xsr: XMLStreamReader2, xsw: XMLStreamWriter2): Unit = {
     xsr.getEventType() match {
       case START_ELEMENT => {
-        xsw.writeStartElement(xsr.getPrefix(), xsr.getLocalName(), xsr.getNamespaceURI())
+        if (xsr.isEmptyElement()) {
+          xsw.writeEmptyElement(xsr.getPrefix(), xsr.getLocalName(), xsr.getNamespaceURI())
+        } else {
+          xsw.writeStartElement(xsr.getPrefix(), xsr.getLocalName(), xsr.getNamespaceURI())
+        }
         for (i <- 0 until xsr.getNamespaceCount()) {
           xsw.writeNamespace(xsr.getNamespacePrefix(i), xsr.getNamespaceURI(i))
         }
@@ -155,8 +159,13 @@ object XMLTextInfoset {
             xsr.getAttributeValue(i)
           )
         }
+        if (xsr.isEmptyElement()) {
+          // skip the next END_ELEMENT event since writeEmptyElement above causes the
+          // XMLStreamWriter to handle closing the empty element
+          xsr.next()
+        }
       }
-      case END_ELEMENT => xsw.writeEndElement()
+      case END_ELEMENT => xsw.writeFullEndElement()
       case CHARACTERS => xsw.writeCharacters(xsr.getText())
       case COMMENT => xsw.writeComment(xsr.getText())
       case CDATA => xsw.writeCData(xsr.getText())
@@ -189,8 +198,10 @@ object XMLTextInfoset {
 
 class XMLTextInfosetInputter(input: java.io.InputStream) extends api.infoset.InfosetInputter {
 
-  private lazy val xsr: XMLStreamReader = {
-    val xsr = XMLTextInfoset.xmlInputFactory.createXMLStreamReader(input)
+  private lazy val xsr: XMLStreamReader2 = {
+    val xsr = XMLTextInfoset.xmlInputFactory
+      .createXMLStreamReader(input)
+      .asInstanceOf[XMLStreamReader2]
 
     // no need for UnparseError here. If the XML syntax is bad, parser catches it before we get here.
     Assert.invariant(xsr.hasNext())
@@ -256,8 +267,9 @@ class XMLTextInfosetInputter(input: java.io.InputStream) extends api.infoset.Inf
     // wrapper tag. We trim the result to remove whitespace that the outputter
     // may have written with pretty mode enabled.
     val sw = new StringWriter()
-    val xsw =
-      XMLTextInfoset.xmlOutputFactory.createXMLStreamWriter(sw, StandardCharsets.UTF_8.toString)
+    val xsw = XMLTextInfoset.xmlOutputFactory
+      .createXMLStreamWriter(sw, StandardCharsets.UTF_8.toString)
+      .asInstanceOf[XMLStreamWriter2]
     xsw.writeStartDocument()
     while (
       xsr.getEventType() != END_ELEMENT || xsr.getLocalName() != XMLTextInfoset.stringAsXml

daffodil-core/src/main/scala/org/apache/daffodil/runtime1/infoset/XMLTextInfosetOutputter.scala

@@ -31,6 +31,9 @@ import org.apache.daffodil.lib.util.Indentable
 import org.apache.daffodil.lib.xml.XMLUtils
 import org.apache.daffodil.runtime1.dpath.NodeInfo
 
+import org.codehaus.stax2.XMLStreamReader2
+import org.codehaus.stax2.XMLStreamWriter2
+
 /**
  * Writes the infoset to a java.io.BufferedWriter as XML text.
  *
@@ -148,11 +151,12 @@ class XMLTextInfosetOutputter private (
     // logic also skips the START_DOCUMENT event so that the XML declaration is
     // not written in the middle of our XML infoset
     val sr = new StringReader(str)
-    val xsr = XMLTextInfoset.xmlInputFactory.createXMLStreamReader(sr)
-    val xsw = XMLTextInfoset.xmlOutputFactory.createXMLStreamWriter(
-      writer,
-      StandardCharsets.UTF_8.toString
-    )
+    val xsr = XMLTextInfoset.xmlInputFactory
+      .createXMLStreamReader(sr)
+      .asInstanceOf[XMLStreamReader2]
+    val xsw = XMLTextInfoset.xmlOutputFactory
+      .createXMLStreamWriter(writer, StandardCharsets.UTF_8.toString)
+      .asInstanceOf[XMLStreamWriter2]
     Assert.invariant(xsr.getEventType() == START_DOCUMENT)
     while (xsr.hasNext()) {
       xsr.next()

daffodil-test/src/test/resources/org/apache/daffodil/infoset/stringAsXml/namespaced/binMessage_01.dat

@@ -9,8 +9,8 @@
   <?processing instruction?>
   <field><![CDATA[Field ]]> with <![CDATA[cdata]]> </field> here is mixed content
   <field>spaces</field>    <field>   spaces   </field> and more mixed content
-  <field/> and more mixed content
-  <field/>
+  <field attr="foo"/> and more mixed content
+  <field attr="bar"></field>
   <field>entity references: &lt; > &amp; " ' ©</field>
   <field>CR</field>&#xd;<field>LF</field>
 <field>CRLF</field>&#xd;