Please describe the problem:
readerForFile processes an XInclude under a Read() loop.  A
NodeType()==XML_READER_TYPE_ELEMENT is usually followed by a
NodeType()==XML_READER_TYPE_END_ELEMENT, but not when IsEmptyElement() is True.
 However, in the included file, the END_ELEMENT *does* follow IsEmptyElement.

I build a stack of nested element names.  I push to the stack when the element
is opened and pop from the stack when the element is closed.  In between, I look
at the state of the stack to determine whether I have come to an appropriate
place in the document to do some additional processing.  

I don't know when to pop the closed element from the stack if it is an empty
element.

Steps to reproduce:
Here are two input files and a Python script:

o The first file is called welcome.xml.  It contains an XInclude of the second file.

<?xml version="1.0" encoding="ISO-8859-1"?>
<!DOCTYPE chapter PUBLIC "-//OASIS//DTD DocBook XML V4.4//EN"
   "http://www.oasis-open.org/docbook/xml/4.4/docbookx.dtd" >
 
<chapter id="welcome">
  <?dbhtml filename="welcome.html"?>
 
  <title>Welcome to BLFS</title>
 
  <para>The Beyond Linux From Scratch book is designed to carry on from where
  the LFS book leaves off. But unlike the LFS book, it isn't designed to be
  followed straight through.  Reading the <xref linkend="whichsections"/>
  part of this chapter should help guide you through the book.</para>
 
  <xi:include xmlns:xi="http://www.w3.org/2003/XInclude" href="conventions.xml"/>
 
</chapter>

o The second file is called conventions.xml.  Each file contains an example of
an empty element.

<?xml version="1.0" encoding="ISO-8859-1"?>
<!DOCTYPE sect1 PUBLIC "-//OASIS//DTD DocBook XML V4.4//EN"
   "http://www.oasis-open.org/docbook/xml/4.4/docbookx.dtd" >
 
<sect1 id="conventions">
  <?dbhtml filename="conventions.html"?>
 
  <blockquote>
    <para>This form of text is used for several purposes in the
    book but mainly to emphasize important points or to give examples as to
    what to type.</para>
  </blockquote>
 
  <para><ulink url="http://www.linuxfromscratch.org/"/></para>
 
</sect1>

o The Python script is called TestIsEmptyElement.py.

#!/usr/bin/python
 
# TestIsEmptyElement.py
# 2006 Jan 13 . ccr
 
import libxml2
 
ZERO=0
 
__Flags=(
    libxml2.XML_PARSE_XINCLUDE| # Expand xincludes.
    libxml2.XML_PARSE_NOENT|    # Expand entities.  This doesn't look right.
    libxml2.XML_PARSE_NOBLANKS| # Suppress whitespace.
    ZERO)
__Reader=libxml2.readerForFile('welcome.xml','ascii',__Flags)
while __Reader.Read()==1:
    __Type=__Reader.NodeType()
    __Name=__Reader.Name()
    __IsEmpty=__Reader.IsEmptyElement()
    if __Type in [libxml2.XML_READER_TYPE_ELEMENT]:
        print "start %s: IsEmptyElement=%s"%(__Name,__IsEmpty)
    elif __Type in [libxml2.XML_READER_TYPE_END_ELEMENT]:
        print "end   %s"%__Name
 
# Fin


Actual results:
Here is the output from TestIsEmptyElement.py:

start chapter: IsEmptyElement=0
start title: IsEmptyElement=0
end   title
start para: IsEmptyElement=0
start xref: IsEmptyElement=1
end   para
start sect1: IsEmptyElement=0
start blockquote: IsEmptyElement=0
start para: IsEmptyElement=0
end   para
end   blockquote
start para: IsEmptyElement=0
start ulink: IsEmptyElement=1
end   ulink
end   para
end   sect1
end   chapter


Expected results:
The <xref> element in welcome.xml is an empty element.

The <ulink> element in conventions.xml is also an empty element, but it is
follwed by an END_ELEMENT node, whereas <xref> is not.

Does this happen every time?
Yep.

Other information:
Nope.