Re: gnubol: Anyone actually working on Cobol

["Boris Kortiak" <BKortiak@genlyte.com>]

Maybe I'm confusing my levels (remember I'm just a COBOL grunt, never wrote a compiler).

I suspect that it would indeed be something to handle at the lexical level.

The idea would be to have a basic nucleus of the COBOL compiler (and here I mean it in its broadest sense) which would take care of all the strict interpretation of the standard.  Additional functionality (or extensions) should be able to be added through some simple mechanism which would allow development of the additional functions to go on separately and be integrated in a fairly seamless manner.

Anyway these are my thoughts, perhaps Justin could expound from his point of view (since I am tagging onto what he said).

>>> Tim Josling <tej@melbpc.org.au> 11/11/99 03:46PM >>>
Could you expand on this. The parser generators are not very
amenable to modular grammars at all. That is in the sense that
you can easily turn parts of the grammar on and off. It's pretty
easy to turn keyworks on and off via the lexical analyser.

Tim Josling

Boris Kortiak wrote:
> Justin,
> I suspect that while it may be a pain, it would turn out to be the best way in the long run.  Plus it would allow for more "individual" COBOL grammers to exist.  If I want a verb PREFORM to mean the same thing as PERFORM (cause I have real trouble spelling <g>), with a modular approach I could easily (well easier than rewriting the whole thing) add it in.
<snip>


--
This message was sent through the gnu-cobol mailing list.  To remove yourself
from this mailing list, send a message to majordomo@lusars.net with the
words "unsubscribe gnu-cobol" in the message body.  For more information on
the GNU COBOL project, send mail to gnu-cobol-owner@lusars.net.