| 1 |
On Fri, 20 Jul 2012 13:55:46 -0400 |
| 2 |
Mike Gilbert <floppym@g.o> wrote: |
| 3 |
> On Fri, Jul 20, 2012 at 2:54 AM, Ciaran McCreesh |
| 4 |
> <ciaran.mccreesh@××××××××××.com> wrote: |
| 5 |
> > On Thu, 19 Jul 2012 18:34:41 -0400 |
| 6 |
> > Mike Gilbert <floppym@g.o> wrote: |
| 7 |
> >> On Thu, Jul 19, 2012 at 5:13 PM, Zac Medico <zmedico@g.o> |
| 8 |
> >> wrote: |
| 9 |
> >> > On 07/19/2012 06:14 AM, Ralph Sennhauser wrote: |
| 10 |
> >> >> Could be that Portage re-exports a sanitized |
| 11 |
> >> >> LINGUAS tough, but I doubt it. |
| 12 |
> >> > |
| 13 |
> >> > Portage does sanitize it if there are any linguas_* flags in |
| 14 |
> >> > IUSE, otherwise it lets the variable pass through without |
| 15 |
> >> > sanitizing it. |
| 16 |
> >> |
| 17 |
> >> That's good; we definitely don't want to "sanitize" it if there |
| 18 |
> >> are no linuguas_* flags in IUSE. This would break LINUGUAS support |
| 19 |
> >> for many autotools/gettext based packages, where the autotools |
| 20 |
> >> code parses LINGUAS directly and the ebuild does nothing with it. |
| 21 |
> > |
| 22 |
> > If there aren't any linguas_* flags in IUSE, LINGUAS should be |
| 23 |
> > empty, and will be in future EAPIs. Without that, USE dependencies |
| 24 |
> > on USE_EXPAND variables don't work. |
| 25 |
> |
| 26 |
> Do you mean that LINGUAS will be empty, or unset (undefined) in an |
| 27 |
> ebuild context? The difference is significant here. |
| 28 |
|
| 29 |
For EAPIs before 5, LINGUAS contains *at least* the things in IUSE |
| 30 |
intersected with the ones the user has enabled, with the linguas_ |
| 31 |
stripped. It's not just "the environment variable in make.conf", since a |
| 32 |
user might put linguas_en in package.use. |
| 33 |
|
| 34 |
For EAPIs 5 and onwards, LINGUAS contains only those things, and |
| 35 |
definitely won't contain anything else. |
| 36 |
|
| 37 |
-- |
| 38 |
Ciaran McCreesh |
Archives