XML document --> lightweight in-memory tree --> XDM --> Process! ... Who created the XDM data model: the XML Processor or the XSLT Processor?

Hi Folks,

Thanks for the excellent replies yesterday. 

I learned a lot.

To recap, an XML document is ingested by an XML processor, which builds a lightweight in-memory tree of it, or it generates a sequence of SAX events:

XML document --> lightweight, in-memory tree

or 

XML document --> sequence of SAX events

Good.

Now sitting over here is an XML application -- an XSLT processor -- that is eager to get that XML document.

However, the XSLT processor wants the XML document in particular form; it wants it in the form specified by the XDM data model.

So here's what we've got:

XML document --> lightweight, in-memory tree --> XDM --> XSLT processing

or 

XML document --> sequence of SAX events --> XSD --> XSLT processing

How do we get from the lightweight, in-memory tree to XDM? From the sequence of SAX events to XDM?

Specifically, who creates the XDM data model? 

Does the XML processor create the XDM and then hand it over to the XSLT processor?

Or does the XSLT processor create the XDM?

Next, to be sure that I see the pattern, let's switch over to another XML application:  XML Schema validators.

XML Schema validators don't operate on the XDM data model; instead, they operate on the XML Infoset.

Same question as above: who creates the XML Infoset? The XML processor or the XML Schema validator?

/Roger