Appended below is the information that I got from the StarOffice/OpenOffice engineering.
-- --- Gal Shalif. Staff Engineer. Engineering, MDE Israel [EMAIL PROTECTED] Office at Israel: +972-9-9711256 x/12256 Mobile: +972 55 243285 Fax: +972-9-9711222
For the spellchecker implementation please have a look at http://sw.openoffice.org/drafts/linguistic_howto.html everything still applies but lately (in newest versions) registration of the component may be needed in a different file then the applicat.rdb.
The whole set of interfaces to be implemented is ::com::sun::star::linguistic2::XSpellChecker, (implies ::com::sun::star::linguistic2::XSupportedLocales) ::com::sun::star::linguistic2::XLinguServiceEventBroadcaster, ::com::sun::star::lang::XInitialization, ::com::sun::star::lang::XComponent, ::com::sun::star::lang::XServiceInfo, ::com::sun::star::lang::XServiceDisplayName but if you modify the sample stated in the above URL the task should be quite simple since there are only a few modifications that need to be done.
For what is expected for external source code you may look at http://external.openoffice.org/ though I'm not aware if that documentation is still up to date.
Gal Shalif - Sun Israel - Software Engineer wrote:
Hello Nadav,
if any of you OpenOffice guys could look into OpenOffice's actual spellchecker (and not some two-year-old documentation), learn what it can do out-of-the-box, and be prepared to answer questions about it, it would be nice.
The StarDivision (Sun subsidiary) are the main body of development for StarOffice/OpenOffice code-base.
I will contact then and ask about a person that can answer technical questions regarding the SpellChecker API.
-- --- Gal Shalif. Staff Engineer. Engineering, MDE Israel [EMAIL PROTECTED] Office at Israel: +972-9-9711256 x/12256 Mobile: +972 55 243285 Fax: +972-9-9711222
Nadav Har'El wrote:
On Tue, Mar 11, 2003, Uri Bruck wrote about "Re: Hebrew SpellChecker and OpenOffice/StarOffice SpellChecker engine":
http://whiteboard.openoffice.org/lingucomponent/dictionary.html
Thanks. This link helped a lot. I hope it's up-to-date (at least in spirit, if not in detail) and that I understood it correctly.
If myspell's affix compression works like I think it does, I belive that getting Hspell to generate a word list in that format is very doable. In fact, suffix compression is very useful for guranteeing a very small memory footprint of the running spellchecker (not only small on-disk footprint, which hspell already has), so Hspell will probably use something like it in the final release of the C interface in any case.
I'll see what I can do about generating a affix-compressed word list in exactly the format needed for myspell, in which case there will be no need for any new C interface for Hspell to work in openoffice! (you'll only be giving up a few minor features of Hspell, like gimatria recognition, and the Hebrew-oriented correction suggestion tricks).
If anybody else wants to take this task upon him or herself (I believe it is an interesting task, and I also think I know quite well how to go about doing it) please contact me in private. Also, if any of you OpenOffice guys could look into OpenOffice's actual spellchecker (and not some two-year-old documentation), learn what it can do out-of-the-box, and be prepared to answer questions about it, it would be nice.
