Re: Why separate Lexical & Parser Generators

hagerman@ece.cmu.edu (John Hagerman)
Fri, 14 Oct 1994 16:09:27 GMT

          From comp.compilers

Related articles
[4 earlier articles]
Re: Why separate Lexical & Parser Generators johnl@cs.indiana.edu (John Lacey) (1994-10-10)
Re: Why separate Lexical & Parser Generators hagerman@ece.cmu.edu (1994-10-10)
Re: Why separate Lexical & Parser Generators wrs@apple.com (Walter Smith) (1994-10-10)
Re: Why separate Lexical & Parser Generators cef@geodesic.com (Charles Fiterman) (1994-10-11)
Re: Why separate Lexical & Parser Generators pardo@cs.washington.edu (1994-10-11)
Re: Why separate Lexical & Parser Generators johnl@cs.indiana.edu (John Lacey) (1994-10-12)
Re: Why separate Lexical & Parser Generators hagerman@ece.cmu.edu (1994-10-14)
Re: Why separate Lexical & Parser Generators adrian@platon.cs.rhbnc.ac.uk (1994-10-21)
Re: Why separate Lexical & Parser Generators hbaker@netcom.com (1994-10-22)
| List of all articles for this month |

Newsgroups: comp.compilers
From: hagerman@ece.cmu.edu (John Hagerman)
Keywords: design
Organization: Carnegie Mellon University
References: 94-10-028 94-10-098
Date: Fri, 14 Oct 1994 16:09:27 GMT

I wrote:
> But when doing *research* it is nice to (1) have a [production-
> class language], and (2) still be able to extend the language.


"John Lacey" <johnl@cs.indiana.edu> writes:
> There are languages for just such an occasion. Scheme would do
> very nicely indeed, and ML or Common LISP are fine alternatives.


Agreed. But I didn't have a choice; I was essentially told "Use this
language." Thus my argument is that production languages *should* be
built to permit extension (through free-ranging comments, in terms of
this discussion) so that they can be used by researchers.


~ John
--


Post a followup to this message

Return to the comp.compilers page.
Search the comp.compilers archives again.