XML Pull Parser
Last published: October 2, 2007 by 'antony'
Defines 10 Classes
Extends 11 Classes
An XML pull parser, with methods designed for easy processing on the basis of tag name tests. Currently ignores namespaces, although they would be trivial to adfd by inverting the test logic for tag matching and putting those methods on String and a new Namespace class.
At the moment the implementation simulates pulling by collecting SAX events in a list - the VW XML parser could operate in pull-mode but I haven't got around to doing it yet. This means that the error behaviour is currently unusual for a pull parser in that parse errors are reported before anything is pulled. It also means that one of the primary benefits of pull parsing over DOM navigation i.e. space efficiency, is lost. Still, fixing that would be transparent to clients.
This package is based on my long experience with a similar development in Java, which was itself a wrapper around XPP3. There is more information at http://www.xmlpull.org/