[prev in list] [next in list] [prev in thread] [next in thread] 

List:       kde-i18n-doc
Subject:    Re: Renaming desktop_<module>_<program>.po file
From:       Burkhard =?ISO-8859-1?Q?L=FCck?= <lueck () hube-lueck ! de>
Date:       2018-04-14 8:56:37
Message-ID: 152380214.1BnGsidFI4 () tuxedo
[Download RAW message or body]

Am Freitag, 13. April 2018, 20:45:26 CEST schrieb Albert Astals Cid:
> El divendres, 13 d'abril de 2018, a les 0:13:52 CEST, Luigi Toscano va
> 
> escriure:
> > Albert Astals Cid ha scritto:
> > > El dilluns, 20 de juny de 2016, a les 23:47:43 CEST, Luigi Toscano va
> > > 
> > > escriure:
> > >> Burkhard Lück ha scritto:
> > >>> Am Dienstag, 12. April 2016, 16:22:03 CEST schrieb Luigi Toscano:
> > >>>> Chusslove Illich ha scritto:
> > >>>>>> [: Luigi Toscano :]
> > >>>>>> So the question is: are there any major blocker for implementing
> > >>>>>> the
> > >>>>>> change from the current pattern to <program>.desktop.po?
> > >>>>> 
> > >>>>> I have just one little note: there already exist one catalog named
> > >>>>> plasma_shell_org.kde.plasma.desktop.po, that is not a desktop PO.
> > >>>>> Not
> > >>>>> sure
> > >>>>> if that will cause some confusion somewhere. And renaming this one
> > >>>>> catalog
> > >>>>> to something else may also be not nice, because it follows the
> > >>>>> reverse-
> > >>>>> domain naming pattern.
> > >>> 
> > >>> All catalogs in trunk kf5 with naming scheme "*desktop.po" but no
> > >>> desktop
> > >>> file:
> > >>> 
> > >>> docmessages: kcontrol_desktop.po + plasma-desktop.po
> > >>> 
> > >>> gui messages:
> > >>> kcm_kwindesktop.po, plasma_applet_org.kde.plasma.showdesktop.po,
> > >>> plasma_runner_plasma-desktop.po, 
> > >>> plasma_shell_org.kde.plasma.desktop.po
> > >>> 
> > >>>>> (In the summit I didn't want to think about this, so as a precaution
> > >>>>> I
> > >>>>> set
> > >>>>> it to be renamed to plasma_shell_org.kde.plasma__desktop.po.)
> > >>>> 
> > >>>> Good point: this won't probably happen too much, but maybe we can
> > >>>> define
> > >>>> a
> > >>>> custom prefix which is not going to conflict with normal use cases?
> > >>>> Maybe
> > >>>> <program>._desktop.po
> > >>>> <program>._json.po
> > >>>> ?
> > >>> 
> > >>> What these catalogs (desktop, json, appdata, mimetypes) have in common
> > >>> is
> > >>> that they are generated by scripty for internal use only.
> > >>> 
> > >>> We already have the suffix "_qt.po" to distinguish between Gettext and
> > >>> Qt
> > >>> translation system.
> > >>> 
> > >>> Similar we could use a suffix like e.g. "_scripty.po" for the catalogs
> > >>> generated by scripty for internal usage.
> > >>> 
> > >>> *_desktop_scripty.po
> > >>> *_json_scripty.po
> > >>> *_mimetypes_scripty.po
> > >>> *_appdata_scripty.po
> > >>> or a reverse naming scheme
> > >> 
> > >> Using scripty as marker is a possible solution. I personally don't like
> > >> too
> > >> much to use "scripty" explicitly but I wouldn't oppose the solution.
> > >> I'd like to propose to break the _<something>_scripty.po part with a
> > >> dot
> > >> instead of a simply _, but it would be just for clarity.
> > > 
> > > Anything that is unambiguous enough that it won't be part of a "regular"
> > > catalog is good enough
> > 
> > Resurrecting this 2-years old thread because I think that the only blocker
> > to the rename is now fixed.
> > 
> > Any final opinion about the pattern?
> > 
> > Burkhard's proposal was:
> > *_desktop_scripty.po
> > *_json_scripty.po
> > *_mimetypes_scripty.po
> > *_appdata_scripty.po
> > 
> > I suspect that a dot is needed before the first _:
> > file._desktop_scripty.po
> > 
> > As I mentioned, I'd prefer to remove scripty and use a pattern like
> > foo._<type>_.po or foo.__<type>__.po.
> > 
> > What do you think?
> 
> I really don't care :D
> 
Same for me, we just need an arbitrary pattern to distinguish between regular 
catalogs and catalogs generated by scripty for internal use.

-- 
Burkhard Lück

[prev in list] [next in list] [prev in thread] [next in thread]