hash table to replace Judy
gyepi at praxis-sw.com
Wed Sep 18 18:37:13 EDT 2002
On Tue, Sep 17, 2002 at 11:58:39PM -0400, Doug Beardsley wrote:
> Hmmm, before bogofilter came out, I wrote my own implementation of Paul
> Graham's idea, except I did my own word dictionary/hash table/search
> tree. I was originally frustrated that bogofilter required the Judy
> library (and more frustrated when dependence on BerkeleyDB came in). I
> figured that it would be better to just use Judy and forget about
> proprietary implementations. Since we are talking about using new code
> for that job, I thought I would throw in my own attempt. I won't
> include it in this message, but if you would like to see it, I will be
> happy to send it to you or post it to this list. What are some other
> opinions about this? Stick with Judy or find something new to eliminate
> the dependency?
By all means, as I just mentioned in another email, we should consider all alternatives before things
jell too much. Reducing dependencies on external libraries may be a good option. Especially since the
code can then be tuned for the task. Can you send me a copy please?
My benchmark stuff is available at: http://www.praxis-sw.com/software/wordfreq_benchmark.tar.gz
There is no documentation, save the source, but it is pretty straight forward.
$ make bench
More information about the Bogofilter-dev