Text Search Engine that works with Python

[Tim Bell]

Fernando Rodr?uez <frr at ThouShallNotSpam.EasyJob.NET> wrote in message news:<mt2h8us6sl06q3v47jcrkq5lvdcqn7r7ta at 4ax.com>...
> On 7 Mar 2002 18:41:45 -0800, writeson at earthlink.net (Doug Farrell) wrote:
> 
> >is because of all the other media that goes on the CD. Our current
> >search engine is a piece of junk, but works in the C++ environment of
> >our current application, which is only for Windows. I am considering

> Whatever you use, you will probably have to write the Python interface. 
> If I was in your situation I  would write a Python interface to the mg system
> that someone already recommended to you...

MG has already been adapted for use on multimedia CD ROMS as the Greenstone
Digital Library software <http://www.greenstone.org/>.  Most of the work was
done at the University of Waikato in Hamilton, NZ, and I'm sure it would be
worth your while investigating their approach.  (They don't use Python
anywhere, to my knowledge.)

While the Greenstone system is cross-platform (including even Windows 3.1),
there is considerable effort required in getting it to work on each platform.
Clearly a language such as Python (coupled with a suitable GUI toolkit) would
be very useful here.

Also, if you start getting into implementing a new text indexing and
compression system, or even just building a Python API for MG, I strongly
recommend you read Managing Gigabytes, 2nd Edition, by Witten, Moffat and
Bell; Morgan Kaufmann, 1999.  (My name is similar to the third author's,
but we're different people.)  There's more info about the book here:
<http://www.cs.mu.oz.au/mg/>.  Much of the work they describe has been
implemented in the MG system.

Tim.
-- 
Tim Bell - bhat at cs.mu.oz.au - Dept of Comp Sci & SE - Uni of Melbourne, Aust.