Mike Mason wrote
>I'm sure one of the IBM developers said that XML4C internally uses the
>SAX interface to build a tree for final DOM exposure. This would seem to
>be sensible reuse of a SAX parser, although it might raise efficiency
>questions.
The C++ parser has internal interfaces that are similar to SAX in
concept and flavor, but extended to provide more detailed information
about the document being parsed.
Look at the source files in the directoryy xml-xerces/c/src/framework for
the details. XMLDocumentHandler.hpp and XMLEntityHandler.hpp are good
places to start.
Both the SAX and DOM parsers are implemented on top of these framework
interfaces. SAX is a very thin, almost trivial, wrappering. The DOM
parser installs handlers that build the DOM tree. Both are really pretty
simple. Look in the directory .../src/parsers
-- Andy
