Re: flex code memory usage

Chris Isbell <ChrisIsbell@voila.fr>
8 Feb 2004 22:02:57 -0500

          From comp.compilers
Related articles
flex code memory usage stefan.fruehwirth@gmx.at (=?ISO-8859-1?Q?Stefan_Fr=FChwirth?=) (2004-02-01)
RE: flex code memory usage rajaram@acmet.com (RERA) (2004-02-04)
Re: flex code memory usage stefan.fruehwirth@gmx.at (=?ISO-8859-1?Q?Stefan_Fr=FChwirth?=) (2004-02-04)
Re: flex code memory usage ChrisIsbell_nospaam@voila.fr (Chris Isbell) (2004-02-04)
Re: flex code memory usage ChrisIsbell@voila.fr (Chris Isbell) (2004-02-08)
| List of all articles for this month |
From: Chris Isbell <ChrisIsbell@voila.fr>
Newsgroups: comp.compilers
Date: 8 Feb 2004 22:02:57 -0500
Organization: Private individual
References: 04-02-054
Keywords: lex
Posted-Date: 08 Feb 2004 22:02:57 EST

On 4 Feb 2004 21:43:40 -0500, "RERA" <rajaram@acmet.com> wrote:


>Using flex for a embedded system :O. I don't think it's a better
>idea. Is your embedded product that complex? What does your chip
>support ?
>
>Hand written lexer/parser will be a better idea.


That has not been my experience. My flex and bison code comes out at
about 26k bytes of ROM and very little RAM. The performance is good
and the code is very robust. It has replaced a hand-written parser and
is better than it in almost all respects - especially flexibility and
maintainability. I would not wish to go back to the hand-written
parser.


My processor is a 16-bit microcontroller (Infineon C167) running at
20MHz, so it is not exactly a powerful system by today's standards. In
addition to the parser it is running a TCP/IP stack with an Ethernet
interface.
--
Chris Isbell
Southampton
UK
[Don't forget that Lex and yacc were originally written on a PDP-11 with
a 64K byte user address space running at about 1MHz. -John]


Post a followup to this message

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