Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!csd4.milw.wisc.edu!lll-winken!uunet!mcvax!kth!draken!liuida!mikpa
From: mikpa@massormetrix.ida.liu.se (Mikael Patel)
Newsgroups: comp.lang.forth
Subject: Re: FIFTH?
Message-ID: <1245@massormetrix.ida.liu.se>
Date: 14 Apr 89 12:45:25 GMT
References: <8904140055.AA18922@jade.berkeley.edu>
Organization: CIS Dept, Univ of Linkoping, Sweden
Lines: 41
In-reply-to: HALDANE@UFPINE.BITNET's message of 13 Apr 89 17:49:00 GMT

Its quite easy to simulate operator overloading in Forth using 
vocabularies. In my C based Forth 83 implementation, XFORTH, I have added
a facility in the Forth top-loop so that overloading and casting ala C are
possible. 

For instance, in a vocabulary with string function one may have a function
plus, "+", for string concatination (not $+ as often advocated).

If the vocabulary is in the search order and found first there's no
real problem. To hard bind to the plus operator from this vocabulary in
other cases when, for instance, forth preceeds the string vocabulary
in the search order, I allow a "casting" operation: "(string) +"

The implementation is very simple (as everything in Forth). In the
top loop, normally interpret, if "find" fails I examine the token
it see if it could be a casting operation, i.e., some word in
parenpheses. In this case I remove the parenpheses and call "find" again.

If it returns with a vocabulary entry I just scan the next symbol and 
repeat the procedure until it either fails, or an entry is found
(which is not a vocabulary). So, "(float) +", would mean plus from float. 

Now all of this talk is early binding. If you want to do late binding ala
Smalltalk the problem is somewhat different. Maybe we can discuss late
binding in Forth in this news group because there alot of way of introducing
it. So let's continue this discussion on late binding!!

May the Forth be with you...

Mikael R.K. Patel
Researcher and Lecturer
Computer Aided Design Laboratory
Department of Computer and Information Science
Linkoping University
S-581 83  LINKOPING
SWEDEN

Phone: +46 13281821
Telex: 8155076 LIUIDA S			Telefax: +46 13142231
Internet: mip@ida.liu.se		UUCP: ...!enea!liuida!mip
Bitnet: MIP@SELIUIDA			SUNET: LIUIDA::MIP