Re: Modification of the spell checker

[Thorsten Behrens <thorsten.behrens@sun.com>]

Hi Jan,

yesterday, you wrote:

> I decide to solve this problem and I got it by using Metric Access
> Methods which are suitable for similarity information retrieval.
>
Could you elaborate? Do you have a specific query method in mind?

> Our method is quite better than the existing one proposed by OpenOffice
> and it presents a logarithmic time to retrieve the similar words. Right
> now, we would like to contribute and add this data structure to the open
> office but I do not have idea about the structure of OO code. May you
> help me to do that?.
>
I'm only vaguely familiar with the spell checker internals, but AFAIK, OOo
itself is rather agnostic about the exact algorithm used to come up with
similar words - that's an implementation detail of the checker engine
used. For OOo, that would be aspell, for SO, one out of several commercial
products.

Note: your question might reach a wider audience on the OOo developer
lists (dev@lingucomponent.openoffice.org and dev@sw.openoffice.org, for
that matter).

HTH,

-- 

Thorsten

If you're not failing some of the time, you're not trying hard enough.