Created attachment 513518 [details, diff]
app-office/scribus should be migrated from LINGUAS to the L10N USE_EXPAND variable, proposed patch is attached.
Some complications are:
- Language tags like cs_CZ or sv_SE are considered duplicates of their respective generic tags (i.e., cs and sv), when the territory doesn't add any useful information. Therefore we cannot add all items from IUSE_LINGUAS as L10N values.
- I have added a new function map_lang() that will map the duplicates to the generic tags. Valid region subtags like pt_PT and pt_BR will be retained, of course.
- Note that map_lang() operates in the reverse direction, as compared to the previous translate_lang().
- IUSE will contain some duplicate entries, e.g. two times "es". This is ugly but shouldn't harm. (Eliminating them will presumably be quadratic in time, and this has to run in global scope.)
The following values would have to be added to l10n.desc:
de-CH - German (Switzerland)
de-DE - German (Germany)
so - Somali
Created attachment 513860 [details, diff]
Updated patch. This gets rid of the duplicates in IUSE:
> - IUSE will contain some duplicate entries, e.g. two times "es". This is
> ugly but shouldn't harm. (Eliminating them will presumably be quadratic in
> time, and this has to run in global scope.)
Tested with 1.4.6 and 1.5.3-r1. Both build fine here, and install the documentation and localisation files (in the case of 1.5.3-r1) as asked for by L10N.
The bug has been closed via the following commit(s):
Author: Ulrich Müller <email@example.com>
AuthorDate: 2018-01-09 10:12:18 +0000
Commit: Ulrich Müller <firstname.lastname@example.org>
CommitDate: 2018-01-09 13:37:36 +0000
app-office/scribus: Migrate from LINGUAS to L10N.
Package-Manager: Portage-2.3.19, Repoman-2.3.6
app-office/scribus/scribus-1.4.6.ebuild | 47 +++++++++++++++++++-----------
app-office/scribus/scribus-1.5.3-r1.ebuild | 44 ++++++++++++++++++----------
app-office/scribus/scribus-9999.ebuild | 44 ++++++++++++++++++----------
3 files changed, 86 insertions(+), 49 deletions(-)
Sorry for rushing this. We were getting a lot of pressure from users unhappy with the intermediate state of the transition (i.e. linguas_* defined as normal USE flags but no longer expanded from the LINGUAS variable). See for example bug 643598, or the forum thread at https://forums.gentoo.org/viewtopic-t-1074924.html.
(In reply to Ulrich Müller from comment #3)
> Sorry for rushing this.
on the contrary, thanks you for your work and providing the patch.