home
***
CD-ROM
|
disk
|
FTP
|
other
***
search
/
ftp.cs.arizona.edu
/
ftp.cs.arizona.edu.tar
/
ftp.cs.arizona.edu
/
icon
/
newsgrp
/
group01b.txt
/
000191_icon-group-sender_Wed Dec 12 16:33:13 2001.msg
< prev
Wrap
Internet Message Format
|
2002-01-03
|
1KB
Return-Path: <icon-group-sender>
Received: (from root@localhost)
by baskerville.CS.Arizona.EDU (8.11.1/8.11.1) id fBCNWv611359
for icon-group-addresses; Wed, 12 Dec 2001 16:32:57 -0700 (MST)
Message-Id: <200112122332.fBCNWv611359@baskerville.CS.Arizona.EDU>
From: eka@corp.cirrus.com (Eka Laiman)
Subject: Re: Icon vs Prolog in parsing
To: icon-group@cs.arizona.edu
Date: Wed, 12 Dec 2001 12:09:27 -0800 (PST)
Errors-To: icon-group-errors@cs.arizona.edu
Status: RO
Content-Length: 682
Michele Paterniti wrote:
> I am a Prolog programmer, almost new to Icon (which a find by
> accident, reading a book about languages).I'm tryng to figure out
> which are the advantages of each of them in parsing.
> ...............
The advantage of PROLOG over ICON in parsing is: PROLOG provides the
back-tracking mechanism automatically - it comes free. Therefore you
can describe your parser by simply typing your grammar.
On the other hand, if you use "recursive descent method" to write your
parser, your parser will pretty much mimic your grammar and hence is
not any more difficult than writing it in PROLOG. I have written many
parsers using ICON using this method.
-eka-