On Tue, 14 Dec 1999, Jerome ALET wrote:
> I think a better way should be to have conversion tables for every
> language which allow:
...
> the problem is that the locales SHOULD include this sort of conversion
> tables, and functions to deal with them, but AFAIK it's not the case, so
> we have to made them by ourselves.
>
> what I don't know is what table size we really need and what to put in it.
> I think for ISO_8859_1 the 256 first characters suffice, but I don't know
> for others.
Yes, this has been brought up a number of times (and we haven't
capitalized on it yet). The simplest way IMHO would be a fuzzy algorithm
that would read in user-defined files to serve as these tables and
generate alternate words. Or perhaps htfuzzy would generate a auxilliary
database for faster lookup, I don't know.
A thread came up recently in the context of romanized Japanese with the
same concept.
-Geoff Hutchison
Williams Students Online
http://wso.williams.edu/
------------------------------------
To unsubscribe from the htdig3-dev mailing list, send a message to
[EMAIL PROTECTED]
You will receive a message to confirm this.
