Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!tut.cis.ohio-state.edu!ucbvax!bloom-beacon!athena.mit.edu!chekmate
From: chekmate@athena.mit.edu (Adam Kao)
Newsgroups: comp.lang.forth
Subject: Re:  writing a C interpreter in Forth?
Message-ID: <1990Jan9.125031.232@athena.mit.edu>
Date: 9 Jan 90 12:50:31 GMT
References: <9001070938.AA07304@jade.berkeley.edu> <1017@acf5.NYU.EDU> <2F._Fxds13@ficc.uu.net>
Sender: news@athena.mit.edu (News system)
Reply-To: chekmate@athena.mit.edu (Adam Kao)
Organization: Massachusetts Institute of Technology
Lines: 25

I've been thinking more about my original proposal; I need to study
the Dragon Book more closely before I can get more specific.  However,
I do feel that just reimplementing the C yacc table walker in Forth
makes for a more traditional parser architecture than I had in mind.
I was hoping for the easy extensibility that Peter mentioned.  I need
to study the Dragon Book to make sure I understand these and other
options and the tradeoffs.  Maybe I can even think of an option not in
the Dragon Book (well, I can dream, can't I?).

On a separate topic, I'm trying to decide which Forth to use.  I've
played with JForth and I like it a lot, but since this is a Forth
extension (not an application) I hesitate to depend on features
available from only one vendor on one machine (eg JForth's ODE).  It
might be better to use some standard UNIX Forth (memory protection!),
possibly with redistributable extensions like those in Dick Pountain's
book (those ARE redistributable, right?)  Opinions?  Advice?

Finally, I am very interested in the experiences of the people who
have implemented these kinds of things, and I'd love to get my hands
on any of the code you wrote, if that's possible.  Tayfun Kocaoglu's
thesis sounds like exactly what I should be reading -- could you put
me in touch with him, Mitch?

Thank you for your time,
Adam