Path: utzoo!utgpu!news-server.csri.toronto.edu!mailrus!wuarchive!usc!apple!snorkelwacker!spdcc!ima!esegue!compilers-sender
From: terry@cs.tamu.edu (Terry Escamilla)
Newsgroups: comp.compilers
Subject: Seeking parser generator topics
Message-ID: <9008241906.AA11285@cssun.tamu.edu>
Date: 24 Aug 90 19:06:35 GMT
Sender: compilers-sender@esegue.segue.boston.ma.us
Reply-To: terry@cs.tamu.edu (Terry Escamilla)
Organization: Compilers Central
Lines: 30
Approved: compilers@esegue.segue.boston.ma.us

Keywords; parse, question, yacc

I am interested in designing a parser generator tool using object-oriented
approaches.  For example, one could completely write a parser generator in
Eiffel, regardless of the output language for the parser itself.  The
question, though, is how to come up with an interesting slant to the topic
so that it is sufficient for a dissertation.  Some ideas that I have
include:

- reusable components for creating similar parser generator tools,

- extending the concept of grammar inheritance proposed by
  Aksit, et. al., from University of Twente in the Netherlands, or

- making every component of the grammar an object with multiple attributes
  for grammar symbols (contrary to YACC), thus giving a very fine grain of
  reusability (though I don't know what one would do with this).

Another possibility is to take an OO approach for compiler specifications.
I am unfamiliar with current research in building a compiler which is
specified by its grammar and a formal semantics.  However, there may be
some potential for OOP in this approach as well.

Thanks for any ideas,

Terry Escamilla
terry@cs.tamu.edu
-- 
Send compilers articles to compilers@esegue.segue.boston.ma.us
{ima | spdcc | world}!esegue.  Meta-mail to compilers-request@esegue.