Path: utzoo!attcan!uunet!bionet!LANL.GOV!michael%genome
From: michael%genome@LANL.GOV (Michael J. Cinkosky)
Newsgroups: bionet.molbio.genome-program
Subject: bnf for genbank
Message-ID: <9009242002.AA05503@genome.lanl.gov>
Date: 24 Sep 90 20:02:08 GMT
Sender: daemon@genbank.bio.net
Lines: 39

Tom Schneider writes:

> In article <922@dimebox.cs.utexas.edu> read@cs.utexas.edu (Rob) writes:
> >*  Does anyone have a lexer/parser for the GenBank Feature Table?
> >*  Would anone want one if wrote one?
> 
> This is an excellent idea.  But it depends on a "Definition of Genbank", a
> document that (to my knowledge) not ever made public.  Has this been released?
> Is it complete?  Does it include a complete BNF of the entire data structure
> (not just the features table)?  Does it define allowed values of all parameters
> in the structure?  If these things are not in place and documented, your parser
> is doomed eventually...  (I spent many years attempting as a GenBank Advisor to
> get them to define the database, and I don't think they have.)
> 
> >Robert L. Read				University of Texas
> >read@cs.utexas.edu			Center for High Performance
> >(512)-477-1240				Computation
> 
> Tom Schneider
> National Cancer Institute
> Laboratory of Mathematical Biology
> Frederick, Maryland  21702-1201


Tom is only partly correct.  There is not a bnf for the entire genbank
format, but there is a bnf for the new feature table format ( the bnf is
part of the feature table document).  This is all that is
required in order to write the lexer/parser originally requested.
(However, it should be mentioned that Dan Davison has had trouble
producing a lexical analyser based on that bnf (basically by converting
the bnf into a lex file), and I am not aware of any other attempts.
It is possible that the bnf presented in the feature table description
document contains an error, but no one here has had a chance to chase
down the problem.)

Michael Cinkosky
GenBank Computation Domain Leader
Los Alamos National Laboratory
Los Alamos, NM  87545