Also known as SAX, this is an alternative to the DocumentObjectModel. The DocumentObjectModel loads an entire document into memory, producing a tree of objects which you can traverse as you see fit. SAX, by contrast, requires you to supply an object with certain standard methods, to a parser which trolls through the XML document linearly, and calls your object every time it sees something interesting. Thus, SAX doesn't have to contain the whole XML document in memory.
Contributor: EdwardKiser.
See http://www.saxproject.org/ for details.
SAX can be useful, but for me not that Simple! For simplicity I make assumptions about valid XML structures and then write my own wrapper classes/functions around a DOM. Check out XmlLite. --PaulCaswell