Path: utzoo!mnetor!uunet!husc6!linus!necntc!ima!johnl
From: ejp@ausmelb.oz.au (Esmond Pitt)
Newsgroups: comp.compilers
Subject: Re: Using Gnu's Bison
Message-ID: <918@ima.ISC.COM>
Date: 22 Mar 88 00:57:23 GMT
References: <1145@thorin.cs.unc.edu>
Sender: johnl@ima.ISC.COM
Reply-To: ejp@ausmelb.oz.au (Esmond Pitt)
Organization: Austec International Limited, Melbourne
Lines: 28
Keywords: yacc, lex, bison,
Approved: compilers@ima.UUCP

In article <1145@thorin.cs.unc.edu> turner@unc.cs.unc.edu (Douglass Turner) writes:
>Anyone out there using Gnu's Bison? ...

1. To get yacc's default output names (y.output, y.tab.c and y.tab.h) instead
of Bison's output names, you need to use the -y switch.

2. Bison's generated parser uses the Vax alloca() function, which you may
not have.

3. Despite claims in the code, Bison's generated parser was for me
slightly slower than yacc's, not that this is a genuine concern.

4. Bison itself runs about 7 times as fast as yacc. This is
insignificant on normal grammars, but, on the large (Cobol-85) grammar
I am using, it goes from 15 minutes to 2-1/2 - an obvious boon.

5. Bison's generated parser has a copyright notice in it which prevents
you from embeddding the result in a proprietary product.

6. Apart from the above issues, Bison plugs straight in.
-- 
Esmond Pitt, Austec International Ltd
...!uunet.UU.NET!munnari!ausmelb!ejp,ejp@ausmelb.oz.au
--
Send compilers articles to ima!compilers or, in a pinch, to Levine@YALE.EDU
Plausible paths are { ihnp4 | decvax | cbosgd | harvard | yale | bbn}!ima
Please send responses to the originator of the message -- I cannot forward
mail accidentally sent back to compilers.  Meta-mail to ima!compilers-request