MathGroup Archive 1993

[Date Index] [Thread Index] [Author Index]

Search the Archive

Re: BNF form for MMA?

  • To: mathgroup at yoda.physics.unc.edu
  • Subject: Re: BNF form for MMA?
  • From: Roberto Sierra <73557.2101 at compuserve.com>
  • Date: 24 Jun 93 03:30:56 EDT

Regarding David Harrison's message:
>   I am seeking a Backus-Naur Form or similar formal specification
>   of Mathematica.  Any help would be appreciated.

>   In case you are interested, the reason for this is that we are
>   requiring our undergraduates to do a fair amount of programming
>   in MMA.  I know of a "cheater meter" program that identifies
>   programs that appear to have been copied from each other; it
>   knows BNF plus some specific languages, and I would like to use
>   the tool.  Not that I expect any of *my* students to register
>   on the cheater meter scale, or course, but ...

Offhand, I don't know of the existence of a publicly available BNF
grammar for MMA, though I do remember that a few months ago there
was a similar request from Nick Flann at Utah State University (his
message is attached).

I might be able to suggest a way to simplify the problem, if you're
able to 'whack' your students MMA submissions around a bit within
MMA itself.

The grammar for MMA probably deals mostly with the precedence of
the various prefix/infix/postfix operators, when newlines are
considered significant, and so on.  You might be able to simplify
your parsing requirements significantly by exporting your expressions
using FullForm, InputForm, CForm, etc., and then writing a much smaller
BNF grammar to deal with these simpler formats.

For example, by exporting to FullForm, then you don't have to worry
about operators at all, since they will look like explicit functions,
and all you're left with are elemental constants, identifiers and
function calls.  Alternatively, by using CForm, you might be able to
whack out a suitable subset of the C grammar to accomplish your task.

Hope this helps.


\\|//
 - -
 o o
  J   roberto sierra
  O   tempered microdesigns
 \_/  73557.2101 at compuserve.com


Attached: Earlier message concerning MMA grammars.
=============================================================================
Sender: mathgroup-adm at yoda.physics.unc.edu
Received: from yoda.physics.unc.edu by iha.compuserve.com (5.65/5.930129sam)
        id AA09983; Sun, 2 May 93 04:37:33 -0400
Received: by yoda.physics.unc.edu (4.1/TAS/11-16-88)
        id AA15726; Sun, 2 May 93 00:52:47 EDT
Received: by yoda.physics.unc.edu (4.1/TAS/11-16-88)
        id AA15722; Sun, 2 May 93 00:52:46 EDT
Message-Id: <9305020452.AA15722 at yoda.physics.unc.edu>
To: mathgroup at yoda.physics.unc.edu
Subject: Compiler/Grammar/Parser in Mathematica?
Date: Thu, 29 Apr 93 09:55:06 -0600
From: flann at nick.cs.usu.edu

We are investigating the possibility of writing a partial
compiler for mathematica, and had the following questions:

1) What work has been done in this area already? (I could
   find no references?)

2) Is there a published grammar for mathematica programs
   anywhere?

3) Are there any existing packages that implement parsers in
   Mathematica?


Thanks!

Nick Flann,
Assistant Professor,
Department of Computer Science,
Utah State University,
Logan, UT 84322-4205.

flann at nick.cs.usu.edu






  • Prev by Date: Please post
  • Next by Date: mathematica programming book info
  • Previous by thread: BNF form for MMA?
  • Next by thread: MathTensor Information (Long)