Question about handling DOM parser errors

[Bruce Reid <b....@aristoslogic.com>]

I'm an XML/DOM newbie, and am trying to make sense of the few parsing
examples I've been able to find. One example is in the Xerces-C DOMPrint
sample application, and the other example is in the book C++ XML. The latter
example is a simpler case, whose only error handling is by means of catching
[XML and DOM] exceptions thrown by the parser. The use of the DOMParser in
the Xerces DOMPrint sample application is considerably more complex, and
employs the following measures in addition to the exception catching done in
the first example:

1) It makes use of the parser's getErrorCount method
2) It uses the parser's setErrorHandler method to attach an ErrorHandler to
do some additional black magic

I'm confused about why so many seemingly overlapping methods of error
handling are being used in the DOMPrint example. I would have thought that
simply catching the proper exceptions should be sufficient. My Other sources
of confusion are:

1) The Xerces-C++ API documentation for the ErrorHandler class (Detailed
Description) states that the class is for use by a SAX application. Yet the
class is being used by DOMPrint, which I don't believe is a SAX application.
2) The same documentation says that if the setErrorHandler method is used,
the parser will use that interface instead of throwing an exception. Yet the
DOMPrint example is both calling setErrorHandler AND catching exceptions.

My quest to find the information that will help me become a self-sufficient
DOM programmer has so far been pretty frustrating. I have obtained a few
books on XML and DOM, and have also been searching the web for information.
Still, I'm having difficulty getting adequate information to make me feel
well prepared to properly use of some of the DOM API. I would be very
grateful if someone could help clarify some of the questions I've mentioned
here, and/or refer me to a source of more detailed information, books, or
examples pertaining to the DOM API in general, and the DOMParser in
particular.

Thank you very much for any assistance,
Bruce Reid