Posted by Oli Filth on 06/08/05 10:23

steve said the following on 08/06/2005 02:35:
> "Oli Filth" wrote:
> > steve said the following on 08/06/2005 00:35:
> > >
> > > For maximum flexibility, I strongly recommend you set up
> > your word
> > > list with regex notation. That way, many different versions
> > of a word
> > > can be caught with a single regex.
> > >
> >
> > How would you do that without blocking a whole load of
> > perfectly
> > innocent words at the same time?
> >
>
> With regex, you can make the criteria as loose or as tight as you
> want. e.g. [ab] only allows "a" or "b" to show up, where as "."
> allows any word to show up. So the control is yours.
>

I'm familiar with regular expressions (I use them all the time ;) ). My
question was more about how you could possibly come up with a regex (or
set of regexes) that do what you suggest, without creating a load of
false negatives (i.e. matching words that are in fact innocent).


--
Oli

[Back to original message]