[xslt20][rwapi] Refactoring base 'Node' classes - comments?

[Shane Curcuru]

Once Lionel's done putting the new proposed QName/Namespace context classes 
in the xpath_rwapi tree, I'd like to propose minor refactoring/renaming of 
a few classes - especially the ones named '*Node'.

Although the names are appropriate at the micro scale - each of these 
classes is a conceptual node - I think it's very confusing to have so many 
Nodes in the hierarchy.  I think I can see obvious confusions with the 
org.w3c.dom.Node class; plus it's very difficult for newcomers in the open 
source world to get up to speed with classnames like this.

Proposal: rename the following classes:
org.apache.xpath.rwapi.datamodel.Node --> XPathNode (or XPDMNode, for 
XPath-DataModel, which is where it's really defined)

org.apache.xpath.rwapi.impl.parser.Node --> JJTNode (for JJTree node, since 
this is really just providing the grammar-driven parser the interface to 
build a tree)

org.apache.xpath.rwapi.impl.parser.SimpleNode --> ASTNode (for Abstract 
Syntax Tree node, since this is really the root of the AST tree we hope to 
build for XPath and XSLT in the future)

org.apache.xpath.rwapi.impl.parser.NodeFactory --> ASTNodeFactory (since 
we'll be using a factory model, I want to make sure users don't confuse 
this with any of the other factories we have in Xalan)

org.apache.xpath.rwapi.impl.parser.QName --> QNameNode or ASTQName (to 
differentiate from the dataholding QName in new:org.apache.xml or 
old:org.apache.xml.utils.

Another obvious change that I think [villard, curcuru, santiagopg] have 
already agreed on in principle is to add some of the functionality from 
org.apache.xalan.xsltc.compiler.SyntaxTreeNode into the 
.impl.parser.ASTNode class, since we'll need a bunch of that stuff for 
runtime implementations.

Comments?  I hate to change the datamodel.Node name since it matches the 
XPath spec directly, but it's really going to cause confusion with 
developers later on.

- Shane