s3 methods on S4 objects

I have J Chambers wonderful text ( Software for data Analysis) and I've been
trying
my hand at some very routine S4 OOP development.

One of the things I was trying to do was to create some very basic S4
classes. The first
was simply a class that had a data.frame as part of its representation.

setClass("df",representation(dirframe="data.frame"))

The object basically contains a data.frame that represents a file directory
listing
with a column named filename, size, time, etc.

And then I have methods for doing various things with this object.

I then tried to tackle the problem of coercing this S4 object to a
data.frame. Again just a learning exercise.

The goal would be able to make a call like this

testFrame <- as.data.frame(x)

where x, was an object of class  "df"

If I try to define "as.data.frame" as a S4 method, then I can make it work,
but I then destroy the S3 functionality
of as.data.frame, so that if I were to try to coerce a matrix to a
data.frame it would work.

as.data.frame(new("A"))

as.data.frame(matrix(0, 3, 5))

So, I guess my question is what do I do, write an s3 method for
as.data.frame that takes  a "df" object as a paramter?
The book wasn't exactly clear ( or I'm not that bright), or is there a way
to make the S4 method I wrote "as.data.frame"
call the S3 method if needed?

	[[alternative HTML version deleted]]