Re: [libreoffice-l10n] How to improve Help translation using vocabulary

Hi Stanislav,

thank you for pointing to the issue, it is exactly my problem. I am
currently fighting with the message-of-the-day strings and it would help
me a lot.

I'll add my +1 to the issue
best
Milos

12/13/19 9:04 PM, Stanislav Horáček wrote:
> Hi,
>
> it seems that this is a similar request as I mentioned here few days ago:
> https://github.com/WeblateOrg/weblate/issues/2183
>
> Would it be a solution for you?
>
> Stanislav
>
> pá 13. 12. 2019 v 11:07 odesílatel Milos Sramek <sramek.milos@gmail.com> napsal:
> > Hi,
> >
> > In translation of the Help (and also the Tip-od-the-day strings) one
> > often needs to look up translation of UI strings. There are two options
> > how to do that: open LO and find the string there, and to search for the
> > string directly in Weblate. Both are cumbersome.
> >
> > There seems to be a third option.
> >
> > The LO manuals are translated using OmegaT. There, one can upload UI
> > translation to its vocabulary and OmegaT then highlights matches
> > directly in the translated text - one can choose the right one by mouse
> > click. This is very efficient and it ensures that the translations are
> > correct.
> >
> > I think that something similar can be achieved also in Weblate. I have
> > imported 16000 string pairs in vocabulary (there was no problem with
> > that) and when translating a string, I get a list of suggestions I can
> > choose from. For example, when translating the string
> >
> > /To start temporarily with a fresh user profile, or to restore a
> > non-working %PRODUCTNAME, use////Help > Restart in Safe Mode/.
> >
> > I can find there translation of /Restart in Safe Mode/. I can transfer
> > its translation to the editor by a mouse click.
> >
> > The problem here is that the algorithm behind vocabulary is very simple
> > - it finds occurrences of each word in the translated string in all
> > vocabulary entries. Thus for the string above I get 2000+ suggestions. I
> > can find the right translation by searching in the browser for the
> > source string. This is not optimal, but still better then the two
> > possibilities mentioned above.
> >
> > I thing that the search algorithm can be modified easily so that it
> > returns only useful matches - I would like to create an issue regarding
> > that in Weblate's github. If one uses just a standard vocabulary (a list
> > of translation of single words), no change would be visible.
> >
> > What do you think about this option? Do you also find it useful?
> >
> > Cheers
> >
> > Milos
> >
> >
> > --
> > email & jabber: sramek.milos@gmail.com
> >
> >
> > --
> > To unsubscribe e-mail to: l10n+unsubscribe@global.libreoffice.org
> > Problems? https://www.libreoffice.org/get-help/mailing-lists/how-to-unsubscribe/
> > Posting guidelines + more: https://wiki.documentfoundation.org/Netiquette
> > List archive: https://listarchives.libreoffice.org/global/l10n/
> > Privacy Policy: https://www.documentfoundation.org/privacy


-- 
email & jabber: sramek.milos@gmail.com


-- 
To unsubscribe e-mail to: l10n+unsubscribe@global.libreoffice.org
Problems? https://www.libreoffice.org/get-help/mailing-lists/how-to-unsubscribe/
Posting guidelines + more: https://wiki.documentfoundation.org/Netiquette
List archive: https://listarchives.libreoffice.org/global/l10n/
Privacy Policy: https://www.documentfoundation.org/privacy