On 02/07/15 13:50, Ciaran McCreesh wrote:
On Sat, 07 Feb 2015 13:46:28 -0500
"Anthony G. Basile" <bluen...@gentoo.org> wrote:
I assume this is a problem with $IUSE and its scope, and not with the
code for in_use() per se? Can you help me understand why you can't
just walk through $IUSE and see if there's a matching flag in there?
It's because of eclasses. The metadata value for IUSE is calculated by
combining the value of the IUSE variable between the ebuild and the
eclasses it uses. That combination isn't necessarily done in bash, and
even if it is, it hasn't necessarily been done when you check
it.
Ah, of course! Thanks. Okay that code has got to go.
I still think the class has merit since it avoids repeated code across
multiple packages and ebuilds. I'll work with Luke-jr on making it clean.
--
Anthony G. Basile, Ph.D.
Gentoo Linux Developer [Hardened]
E-Mail : bluen...@gentoo.org
GnuPG FP : 1FED FAD9 D82C 52A5 3BAB DC79 9384 FA6E F52D 4BBA
GnuPG ID : F52D4BBA
