Home/ Questions/Q 8401551
In Process

The Archive Base Latest Questions

Editorial Team
  • 0
Asked: 2026-06-09T21:50:18+00:00

Before I go into the specific details, I tried this before with Oracle’s parser.v2

  • 0

Before I go into the specific details, I tried this before with Oracle’s parser.v2 XSLT libraries and ran into the same problem (see my first question). However, it works fine when I use C#. 

// from the example I see everywhere on the internet
SAXTransformerFactory stf = (SAXTransformerFactory)TransformerFactory.newInstance();

Templates t_cctagstrip = stf.newTemplates(new StreamSource(new FileInputStream("C:\\path\\to\\first.xsl")));
Templates t_cctable = stf.newTemplates(new StreamSource(new FileInputStream("C:\\path\\to\\second.xsl")));

TransformerHandler th1 = stf.newTransformerHandler(t_cctagstrip);
TransformerHandler th2 = stf.newTransformerHandler(t_cctable);

// transform 1 pipes into transform 2
th1.setResult(new SAXResult(th2));

// transform 2 pipes to System.out
th2.setResult(new StreamResult(System.out));

Transformer t = stf.newTransformer();
t.transform(new StreamSource(new FileInputStream("C:\\path\\to\\source.xml")), new SAXResult(th1));

I’m thinking that maybe the problem lies in the fact that the output from the first transform is a fragment with text at the root level? 

text at root level
<tag> ... tags that need extra processing ... </tag>
more text at root level

Is there something different about how C# processes the transforms? It’s not failing or giving me any warnings that I can tell. It’s just not changing anything.

Share

Leave an answer

You must login to add an answer.

Need An Account,

1 Answer

  1. Editorial Team

    You didn’t in your first question mention the actual problems you’re running into, e.g., providing any log messages, etc. so it was impossible to know why your XSLT transformations are failing. But in this second question you clearly reveal why things are failing.

    In general, XSLT transformations require all parsed content to be at very least well-formed XML. If your source document isn’t wrapped with a “root” element (actually called a document element) then it’s not well-formed XML, therefore not by definition XML, and therefore not suitable for XML processing.

    As a solution, if you can wrap the output from the first transform in a document element you could then continue to parse it as XML. You could then strip it off later via a non-XML process.

    There won’t be any difference between Java or C# in this regard. If the output from the first transformation is not well-formed XML then C# shouldn’t be accepting it as XML. I can’t explain why it might be doing so — C#’s XML processor would be non-compliant if it accepted non-well-formed XML. I imagine the reason you’re not getting any output is that the transformation is simply failing due to the non-XML source. In order to find out what’s actually going on you might set an javax.xml.transform.ErrorListener on your Transformer, and in your ErrorListener dump everything to a log.

    (As a bit of a tangent, you also state you’re required to use Oracle’s XSL processor. Really? I’ll assume there must be some features in Oracle’s implementation that aren’t available in Xalan or Saxon, as you’d be tying your code to an implementation rather than using the JAXP API, which is certainly not recommended practice. It’s probably better to debug problems when coding to the API to be sure that they’re not implementation-dependent, since you aren’t able to change the internal code of an implementation and it’s better to only be debugging your own defects, not Oracle’s. But given your example code above seems to be using the JAXP API I’ll assume the problems you have aren’t actually tied to Oracle’s XSLProcessor.)

    • 0