Return-Path: Return-Path: Received: from ted.cs.uidaho.edu by swan.cl.cam.ac.uk with SMTP (PP-6.0) id <16961-0@swan.cl.cam.ac.uk>; Mon, 27 Jan 1992 09:46:43 +0000 Received: by ted.cs.uidaho.edu (16.6/1.34) id AA10581; Mon, 27 Jan 92 01:26:36 -0800 Sender: info-hol-request@edu.uidaho.cs.ted Errors-To: info-hol-request@edu.uidaho.cs.ted Received: from ux6.tfl.dk by ted.cs.uidaho.edu (16.6/1.34) id AA10577; Mon, 27 Jan 92 01:26:31 -0800 Received: by ux6.tfl.dk (/\==/\ Smail3.1.24.1 #24.2) id ; Mon, 27 Jan 92 10:28 MET Message-Id: Date: Mon, 27 Jan 92 10:28 MET From: kimdam@dk.tfl (Kim Dam Petersen) To: info-hol@edu.uidaho.cs.ted Subject: Re: Generic graphs (was: Overloading of constant names) > [ ... text deleted ... ] > > I'm asking this because I would like to use abstract theories but > found the above naming constraint very annoying. For instance, > I want to define a type :graph and, for any G:graph, get its > adjacency relation by saying adj(G). Later I want to talk about > graphs in which edges are weighted, so I want to define another > type :w_graph. But then, to get the adjacency relation of a > weighted graph WG, I would have to say something like w_adj(WG). > instead of simply adj(WG). If I want to consider yet another > new type :cw_graph of connected weighted graphs, I would have to > introduce yet another set of constant names. Is there a way out? > > - Ching Tsun > If you want to make an abstract theory, you may represent the node and edge types by polymorphic types. Hence, you should define the polymorphic type ":(*node,*edge)graph". You may then specify that "G" is a graph having nodes of type ":node" and edges of type ":edge" by the type constraint "G: (node,edge) graph". Using the polymorphic representation you may derive properties that holds for any graph of type ":(*node,*edge)graph". I leave the actual definition of a polymorphic graph to you. I hope this is useful to you. Regards, Kim Dam Petersen