Apache OpenOffice (AOO) Bugzilla – Issue 99896
Spell checker language can't be defined as Valencian
Last modified: 2017-05-20 11:13:42 UTC
Spell checkers uploaded to the extensions web can't use language name Valencian although they are for Valencian. These extensions are: http://extensions.services.openoffice.org/project/val_dict_racv http://extensions.services.openoffice.org/project/val_dict_llit We're using ISO language code "ca" as recommended for Valencian, but there are conflicts since this code is shared with other languages. Subtag "ca-valencia" could be used to differentiate between languages, but this doesn't seem to be supported by OpenOffice. Subtag "ca-valencia" should be added to the list of supported languages with the name Valencian so we could install Catalan and Valencian spell checkers and use them together by their respective names. Or we'd like to hear any other suggestion that help us achieve the same. The Royal Academy of Valencian Culture (Real Acadèmia de Cultura Valenciana, http://racv.es) has developed this spell checker and has distributed CDs and USB pens freely with the spell checker and OpenOffice, but users have difficulties at setting the language since they can't find it, or they are upset at having to choose the language by another name. Moreover, some of our users need to use the Catalan and Valencian spell checkers simultaneously, mainly at universities and academies, but this is impossible right now and it's a major inconvenient for them.
This issue might be related to Issue 99430, Issue 71636 (Fixed for 2.1)
OOo doesn't support subtags and/or dialects yet, this would need an office-wide and ODF file format implementation of RFC 4646. Sorry to say so, but currently there is no way to have multiple dictionaries per language code. AFAIK this is something planned for future releases. @tl: Do you happen to know some workaround? For discussion of Catalan/Valencian back in October 2006 also see the dev@l10n mailing list archive, http://l10n.openoffice.org/servlets/SearchList?list=dev&searchText=Valencian&defaultField=subject
The result of that discussion was to define it as 'ca-va'.
No, it was ca-XV, ca-VA would be Catalan in Vatican ;-) Anyway, that was only for UI localization, not to be used as document attribution because of the private local extension nature of the XV code.
tl->er: About workaround: Nope. Note really. The only existing one is to misuse the locale of another language.
By using a rarely used language code we could make it compatible with most used languages, but still it would be counter-intuitive for people to set the language right. Do you know when OO will implement subtags or anything that helps this issue? Could it be possible to implement a made-up language code as a workaround until there's a better solution?
This issue is very important for the RACV in order to ease the adoption of Openoffice that it's being promoted from his web (http://racv.es) and distributed in CDs at fairs and to visitors as the preferred suite for Valencian users. The RACV is grateful for this office suite and is willing to collaborate in any way possible to solve this issue. We don't know how we could better help so we're asking for proposals on how to proceed from here. I'm allowed to talk in behalf of the RACV while things get ironed out, but they can be contacted directly too.
As said, OOo will have to implement mechanisms to support variants and BCP47 in the entire suite, including file format. If RACV could provide developer resources to work on this they would be warmly welcomed.
Reset assigne to the default "issues@openoffice.apache.org".