🚀 go-pugleaf

I have been struggling with an algebraic spec for a couple of days and
am beginning to wonder if I am trying to do something that is
impossible.  Intuitively, the feeling is similar to that of trying to
describe a context free language using regular expressions.

Generally, are there _known_ fundamental limitations of algebraic
specifications?  Has anyone studied the "classes" of ADTs that can be
specified using various specification approaches?  If so, I would
appreciate a reference (or a short synopsis, if possible).

More specifically, I am trying to specify a tree that has the notion of
a "current node", and operations to change the current node.  The
current node is a headache, mainly because I have different kinds of
constructors -- some that modify the tree structure (adding/deleting
nodes) and others that modify the current node.  My real problem,
though, is with an operation that changes the current node to the
root.

Does anyone konw offhand if that is possible ... and if it is, how?

If you are interested and can offer more help, we can talk via email
about the specifics of the data structure I am working with.
-- 
Jeff Offutt
Department of ISSE, George Mason University, Fairfax VA
email: ofut@gmuvax2.gmu.edu
phone: (703) 993-1654 (messages: 993-1640)