Sender: crough45 AT amc DOT de
Message-Id: <97Aug12.144935gmt+0100.17063@internet01.amc.de>
Date: Tue, 12 Aug 1997 13:51:33 +0100
From: Chris Croughton <crough45 AT amc DOT de>
Mime-Version: 1.0
To: dog AT fcbob DOT demon DOT co DOT uk
Cc: djgpp AT delorie DOT com
Subject: Re: Arbitrary precision class?
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Precedence: bulk

Dan Goodman wrote:

> Does anyone have, or know of, a C++ class and functions that can do
> arbitrary precision arithmetic (both in very large numbers and very
> accurate numbers)? Please e-mail the source code if you have any...

Funny you should say that...

I'm working on a class wrapper for the GNU MultiPrecision library,
which is a C library available from all GNU archives (look for
gmp-x.y.z.tar.gz - at present x.y.z is 2.0.2 for the latest
version, available from ftp://prep.ai.mit.edu/pub/gnu).  It's a
very comprehensive library, supporting arbitrary precision integer,
rational (fractions) and floating point.

My wrapper is nowhere near ready yet, but I hope to have a usable
version in the next week or so.  I'll willingly distribute the
source of that, but you'll have to get gmp yourself, I'm not sending
a 360K file by mail!

Warning: it's a bit of a pig to unpack, because the directories and 
files are Unix names.  A PKZIP version is available by http only as
http://www.keris.demon.co.uk/bin/gmp202.zip (note that you have to
straight to the directory, there's no link from the homepages) which
can be unzipped using PKUNZIP -d.  You may have to edit file make.bat
to remove the /y on the move command near the end.  If you can wait 
a day or so, I'll zip up my version from home which I know compiles
easily...

Chris C