Newsgroups: comp.ai.nat-lang
Path: cantaloupe.srv.cs.cmu.edu!das-news2.harvard.edu!news2.near.net!news.mathworks.com!uunet!spcuna!ritz!merlin
From: merlin@ritz.mordor.com (Niranjan Hira)
Subject: Re: r
References: <B426Nhr.actamed@delphi.com>
Organization: Mordor International BBS
Date: Wed, 8 Mar 1995 01:44:45 GMT
X-Newsreader: TIN [version 1.2 PL2]
Message-ID: <D53M6M.BBM@ritz.mordor.com>
Lines: 28

ActaMed (actamed@delphi.com) wrote:
: I am trying to write a Augmented Transition Network  parser using a
: series of finite state grammars as the legal states/arcs within the grammar.
: The problem is one of verifying a sequence of symbols and determining
: where the symbols were constructed from (i.e., which grammar produced
: the symbols?). For example, given a character string aabcdbcdggeffhij,
: which of a group of finite state machines produced the sequence? Also,
: if anyone knows which newgroups would be more appropriate for this type
: of question, please let me know. I appreciate any and all help. Thanks
: much.
: Ben B.
: <Actamed@delphi.com>
:  
	This might sound quite silly, but have you thought of maintaining 
a table (i.e. a line for every state in the ATN) consisting of:
	1)  Current state
	2)  Current input string 
	and writing it out to a text file or something similar?

	Hope this helps.

	-- Hira
<----------------------------------------------------------------------------->
|  "I don't let myself believe...I know".  -- Carl Gustav Jung                |
|                                                                             |
|  merlin@ritz.mordor.com                                                     |
|  ONE LOST SOUL!!                                                            |
<----------------------------------------------------------------------------->
