ICU isn't translatable through Launchpad

Bug #421917 reported by Michael Terry
8
This bug affects 1 person
Affects Status Importance Assigned to Milestone
icu (Ubuntu)
Won't Fix
Wishlist
Unassigned

Bug Description

The various locale translations (e.g. country and timezone names) are not translated through gettext but through a custom resource method (http://userguide.icu-project.org/locale/resources).

The source translation files are only stored in ICU's source control (http://source.icu-project.org/repos/icu/icu/trunk/source/data/locales/). In the tarball that ICU ships, the locale data has all been globbed together into a binary file (source/data/in/icudt40l.dat).

Not sure the standard method for fixing this. (A) Encourage upstream to use gettext (doesn't seem likely) or (B) download the source data and transform it to gettext as part of the build process so Launchpad can grab the po/pot files?

Revision history for this message
Michael Terry (mterry) wrote :

You can use the following command to open the .dat file, but I'm not sure how to open the .res files yet, but maybe that's a more sane way to get back the source files than downloading them?

icupkg -d . -x \* icudt40l.dat

Michael Terry (mterry)
summary: - Isn't translatable through Launchpad
+ ICU isn't translatable through Launchpad
Revision history for this message
Steven R. Loomis (srl-icu-project) wrote :

ICU's locale data comes from CLDR. You need to work through the CLDR process for any issues in locale data. http://unicode.org/cldr (I am the lead for ICU for C/C++, and a member of the CLDR project.)

Revision history for this message
Steven R. Loomis (srl-icu-project) wrote :

Just more, CLDR is an XML based interchange language for locale data. One would be welcome to write tools to make that translatable through launchpad or any other method.

(By the way, You can use the tool 'derb' to convert from .res to .txt)

Revision history for this message
Michael Terry (mterry) wrote :

Interesting, Steven. Thanks for the note. So it looks like a workaround for Ubuntu contributors wanting to improve translations is to use http://unicode.org/cldr/apps/survey and suggest string changes? (I note that as of this comment, the database is read-only during the runup to CLDR 1.7.)

Revision history for this message
Steven R. Loomis (srl-icu-project) wrote :

No, don't point people directly to the survey tool. Point them to the main CLDR page, or http://cldr.unicode.org/index/bug-reports - if a user is interested, they can request an account and then have access to help with translations. It might be worth coordinating requests so only a small number of people require accounts.

Revision history for this message
Steven R. Loomis (srl-icu-project) wrote :

Should this bug be closed, with a note that folks should contribute corrections to http://unicode.org/cldr ?

Changed in icu (Ubuntu):
status: New → Won't Fix
importance: Undecided → Wishlist
To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.