[jdom-interest] Couple of issues I need help on, (memory, sax vs dom, etc)

[bob mcwhirter]

> 3)  Sax vs Dom:

Here's how it typically works:

	1) sax-parser reads a file/stream.
	2) it spews SAX events.
	3) A handler reads the events, and builds an object-model.

It works like that if you're using Xerces to build a DOM or
if you use SAXBuilder to build a JDOM.  

In both cases, there's a handler somewhere converting events
into a navigable object model.

If you were to use DOMBuilder, the process would be:

	1) sax-parser reads a file/stream.
	2) it spews SAX events.
	3) A handler reads the events, and builds a DOM object-model.
	4) DOMBuilder walks the DOM object-model, and creates a JDOM
	   that mirros the DOM.

So, instead of SAX -> DOM -> JDOM, just go SAX -> JDOM.

If you want to process large files, and can do so in a sequential
mannner, dom4j's[1] ElementHandler interface is darn nice.  You can
process a -subtree- and detatch it for garbage collection.  Somewhat
a hybrid between events and an object-model.

	-bob

[1] http://dom4j.org/