From lrajchel@ANSI.org  Tue Jan 30 18:09:55 2001
Received: from email1.ansi.org (mail.ansi.org [165.254.114.6])
	by dkuug.dk (8.9.2/8.9.2) with ESMTP id SAA58884;
	Tue, 30 Jan 2001 18:09:54 +0100 (CET)
	(envelope-from lrajchel@ANSI.org)
Received: by email1.ansi.org with Internet Mail Service (5.5.2650.21)
	id <D3J3AHDM>; Tue, 30 Jan 2001 12:09:21 -0500
Message-ID: <2F81C8110D55D411882A0020356797B2151B14@email1.ansi.org>
From: Lisa Rajchel <lrajchel@ANSI.org>
To: "'sc22info@dkuug.dk'" <sc22info@dkuug.dk>
Cc: "'Simonsen, Keld'" <keld.simonsen@dkuug.dk>
Subject: FW: Possible revision of ISO/IEC 14977
Date: Tue, 30 Jan 2001 12:09:20 -0500
X-Mailer: Internet Mail Service (5.5.2650.21)

Dear Members -

This message is forwarded to you for your information and comments, if
applicable.

Best regards,

Lisa Rajchel



-----Original Message-----
From: Roger Scowen [mailto:rss@cise.npl.co.uk]
Sent: Wednesday, January 24, 2001 11:47 AM
To: lrajchel@ANSI.org; rss@cise.npl.co.uk
Subject: Possible revision of ISO/IEC 14977




Lisa

I append a note that I would like circulated to SC22 concerning a
possible revision of ISO/IEC 14977 (Extended BNF) in order to identify
the views and interest (or lack of it) of national bodies and workiing
groups.

Please feel free to edit the note, or send it back if major changes are
desirable or required.

Regards

Roger Scowen

--------------

WG19 --- ISO/IEC 14977 Extended BNF --- POSSIBLE REVISION

SUMMARY --- ISO/IEC 14977 (Syntactic metalanguage --- Extended BNF) has
general applicability whenever it is necessary to define the syntax of
a stream of symbols.  Many people will be content to continue using the
syntactic metalanguage with which they are familiar: but 'Extended BNF'
provides a sensible easy choice for anyone who would otherwise be
tempted to develop yet another variant of BNF.

WG19 is responsible for the standard and wonders whether there is any
interest in revising or extending the standard. An enquiry in BSI has
identified some interest among the working groups. This note asks the
same question to the national bodies and other working groups.

Roger Scowen  (WG19 convener, ISO/IEC 14977 Project editor)
9 Birchwood Grove, Hampton, Middlesex
United Kingdom     TW12 3DU
E-mail: roger.scowen@npl.co.uk

OVERVIEW
'Extended BNF' has several options, a simple version can be summarized:

    SYMBOL(S)                  MEANING
  Unquoted words          Non-terminal symbol
    " ... "               Terminal symbol
    ' ... '               Terminal symbol
    ( ... )               Brackets
    [ ... ]               Optional symbols
    { ... }               Symbols repeated zero or more times
       =                  Defining symbol
       ;                  Rule terminator
       |                  Alternative
       ,                  Concatenation
   (* ... *)              Comment

DESIGN FEATURES --- The advantages of 'Extended BNF' include:  
  a) Full definition --- Users need not attempt to define a 
   syntactic metalanguage: it is very hard to obtain general 
   acceptance.
  b) General applicability --- Any character can appear in the 
   language being defined because every terminal symbol is quoted.
  c) Easy to read and understand --- Comments can be interspersed 
   to explain and cross-reference syntax rules, and a non-terminal 
   symbol need not be written as a single word because space and 
   newline characters are ignored, and concatenation is represented 
   explicitly.
  d) Easy to write and e-mail --- No unusual characters or fonts 
   are required, so it is easy to e-mail a syntax. An explicit 
   'end of rule' symbol means that a line wrapping display will 
   not change the meaning of a syntax.
  e) Extensibility --- 'Extended BNF' is designed for use in an 
   ASCII environment, but further extensions are possible because 
   the metalanguage and language symbols need not be disjoint.

EXISTING USAGE --- Several international standards have already adopted
'Extended BNF' in its earlier existence as BS6154. These standards
include:  Modula-2 (ISO/IEC 10514), VDM-SL (ISO/IEC 13817-1), and
Prolog (ISO/IEC 13211-1), PCTE (ISO/IEC 13719), and Z (ISO/IEC FCD 13568).

POSSIBLE REVISIONS

1) Simplify the standard by removing the option of alternative notations 
which enable a syntax to be expressed in 6-bit ISO 646 notation.

2) Extend the notation so that the characters of ISO/IEC 10646 can be
used to express a syntax.

3) Add a further form of comment:
    
     comment = "//", {character}, new line ;

4) Check ISO 6429:1992 and update as necessary.

5) Make other minor changes found to be desirable.

6) Major change - extend the standard (perhaps as new parts) to include
regular expressions, and/or YACC - LALR(1) notation.

ACTION

Inform Roger Scowen if you have any views on a possible revision of
ISO/IEC 14977, or you wish to nominate members to WG19 to consider the
revision.



R S Scowen, Guest worker (roger.scowen at npl.co.uk)
CMSC, National Physical Laboratory, Teddington, Middlesex, UK. TW11 0LW
The contents of the above message are my views alone and should not be
taken as representing the views held by NPL Management Ltd.
