how can I translate XML files (like the following), which contain
XHTML-formatted text in one element, to PDF using Apache FOP ?

 --- books.xml --->
 <BOOK>
   <AUTHOR>Walt Disney</AUTHOR>
   <PRICE>19.90</PRICE>
   <TITLE>&lt;b&gt;Donald Duck - &lt;i&gt;The True Story&lt;/i&gt;
&lt;/b&gt;</TITLE>
 </BOOK>

 <BOOK>
   <AUTHOR>Matt Groening</AUTHOR>
   <PRICE>25.00</PRICE>
   <TITLE>&lt;b&gt;Homer Simpson For President&lt;/b&gt;</TITLE>
 </BOOK>
   ...
 <--- books.xml ---

You can translate the escape XHTML with a  XML pipeline as implemented
by the smallx [1] project at java.net:

<p:pipe name="xhtml-convert"
   xmlns:p="urn:publicid:IDN+smallx.com:pipeline:1.0";
>

<p:subtree select="TITLE">
<p:parse/>
</p:subtree>

<p:pipe name="xhtml-convert"
   xmlns:p="urn:publicid:IDN+smallx.com:pipeline:1.0";
>

<p:on-error>
<p:template>
<TITLE bad="true"><xsl:copy-of select="/*/node()"/></TITLE>
</p:template>
</p:on-error>