Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!uunet!lll-winken!ubvax!fred
From: fred@ubvax.UB.Com (Fred Noon)
Newsgroups: comp.lang.forth
Subject: Re:  writing a C interpreter in Forth?
Message-ID: <25819@ubvax.UB.Com>
Date: 9 Jan 90 20:33:28 GMT
References: <9001070938.AA07304@jade.berkeley.edu> <1017@acf5.NYU.EDU> <2F._Fxds13@ficc.uu.net> <1990Jan9.125031.232@athena.mit.edu>
Reply-To: fred@ubvax.ub.com.UUCP (Fredrik Noon)
Organization: Ungermann-Bass Enterprises
Lines: 15

An advantage of recursive descent over LALR(1) tables is that error
diagnostics, recovery, or just doing things "on the fly" can be made
very straight-forward to implement.  Using tables one has to build
error recovery into the grammar; with LALR that can get a little
tricky.  I suspect that it is this, rather than FORTH's user interface,
which would be responsible for turning a 2-month project into a 6-month
project, particularly if it were someone's first go-round with lex and
yacc.  (Not being a production FORTHer I can't speak from experience,
however.)

======================================================================
Fredrik Noon  (ubvax.ub.com!fred)
5290 Garrison Circle; San Jose, CA 95123; USA       tel (408) 365-7167
or, when toiling away at Ungermann-Bass                 (408) 562-5632
======================================================================