[jdom-interest] newbie question: parsing in xhtml containing mathml

[Laurent Bihanic]

Morten Andersen wrote:
> Well, the task is pretty simple, but I can't get anything working.
> 
> I want to parse in an xhtml document containing mathml with all the 
> entitities defined like alpha and beta. This should then be transformed 
> using xslt into another xml-document.
> The test-xhtml document is shown below:
...

 > This results in this error:
 > /  "org.jdom.IllegalTargetException: The target "IS10744:arch" is not
 > legal for JDOM/XML Processing Instructions: Processing instruction
 > targets cannot contain colons."

As explained in org.jdom.Verifier.java, JDOM is compliant with section 6 of 
the XML namespace specification 
(http://www.w3.org/TR/REC-xml-names/#Conformance) which states:
"No entity names, PI targets, or notation names contain any colons"

> /Then I tryed to trick the SAXBuilder so that the DTD's are not used by 
> setting the entityResolver to an entityResolver, that doesn't do anything.
>  ---
>         /SAXBuilder builder = new SAXBuilder();
>         builder.setEntityResolver(new NoOpEntityResolver());
> /---
> This results in some output to System.err:
> /        "[Fatal Error] :1:66: White spaces are required between 
> publicId and systemId."
> /But the transformation seems to occur.

This is caused by 2 bugs in JDOMSource: 1. the way the DTD reference is build 
is incorrect (indeed, a white space is missing); 2. No SAX ErrorHandler is set 
to correctly process the error.
Except that DTD information is lost to the XSLT processor, the transformation 
is performed.

Laurent