| Summary: | KRunner lost functionalities due to translating two words into one word | ||
|---|---|---|---|
| Product: | [Plasma] krunner | Reporter: | Martin Flöser <mgraesslin> |
| Component: | general | Assignee: | Plasma Bugs List <plasma-bugs> |
| Status: | RESOLVED FIXED | ||
| Severity: | normal | CC: | aacid, leviatan1, wilderkde |
| Priority: | HI | ||
| Version: | unspecified | ||
| Target Milestone: | --- | ||
| Platform: | Ubuntu | ||
| OS: | Unspecified | ||
| Latest Commit: | Version Fixed In: | ||
Well, IMHO it is more a problem of krunner in general. Albert? Man in which mind is this a translation bug? That's a krunner bug for sure, it's not viable to hardcode the number of words a translation must have Err, sorry, reassigned to wrong product, reassigning to plasma I have the problem with spanish. For example the command 'power profile', in spanish 'perfil de energia' doesn't works. The problem is in kde 4.1, 4.2.4 and 4.3. SVN commit 1073303 by jacopods: Correctly parse strings for non-english locales as well. Thanks to the l10n team for allowing this fix Bonus fix: also make parsing case insensitive BUG: 201243 BUG: 208962 CCMAIL: kde-i18n-doc@kde.org M +44 -20 PowerDevilRunner.cpp M +1 -0 PowerDevilRunner.h WebSVN link: http://websvn.kde.org/?view=rev&revision=1073303 SVN commit 1073306 by jacopods: Fix parsing errors in non-english locale + case insensitive match CCBUG: 201243 CCBUG: 208962 M +44 -20 PowerDevilRunner.cpp M +1 -0 PowerDevilRunner.h WebSVN link: http://websvn.kde.org/?view=rev&revision=1073306 |
Version: (using KDE 4.2.96) Installed from: Ubuntu Packages Reporting in English although it's for German locale as I think the problem might be a general problem. So when you think it is not a problem of the translation please reassign to appropriate product - I guess plasma. I had a look at some KRunners today and I noticed that the KRunner keyword "power profile" is translated to one word "Energieprofil" in German. That's in general OK and the keyword is marked with "Note this is a KRunner keyword". Now I had a look at the code and noticed that there is functionality relying on the fact that the keyword consists of two words (kdebase/workspace/plasma/runners/powerdevil/PowerDevilRunner.cpp line 186 following): if (term.startsWith(i18nc( "Note this is a KRunner keyword", "power profile"))) { foreach( const QString &profile, m_availableProfiles ) { if ( term.split( ' ' ).count() == 3 ) { if ( !profile.startsWith( term.split( ' ' ).at( 2 ) ) ) { continue; } } So as soon as you start to enter another word it will only list those matching. E.g. "power profile Perfor..." will only list the Performance entry. If I want to do the same in German with "Energieprofil Perf..." the non-matching entries are not hidden as the first check for three parts fail. If I add another word the match works again. E.g: "Energieprofil foo Perf..." This is rather annoying as the translation removes functionality. Of course the translators cannot know that there is functionality checking for the number of words. Btw. if the translation would be more words the result would be worse.