-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA256
Hey all -- so with Zac's help I've got a script now that will
significantly speed up testing of EAPI=4-slot-abi from my ('axs') overlay.
The script upgrades your system's portage db (/var/db/pkg) so that any
EAPI=4 atoms currently installed will be upgraded to their
EAPI=4-slot-abi equivalents in the overlay (if they exist). This is
relatively safe as 4-slot-abi is a direct superset of EAPI=4 and so
pkg_{pre,post}rm should not be negatively affected.
The script is bash, using grep, sed and awk heavily; it's not pretty,
but it works. The tricky bit is re-working *DEPEND entries, as they
first need to reflect the updated ebuild (the vast majority of these
changes is the addition of slots and/or slot-operators; i'm trying to
keep all other deps the same as what's in the tree's version of the
ebuild), and then determine each dependency's new SLOT (and sub-slot)
so that it can be substituted into the slot-operator part of the
*DEPEND atom. If I knew python well enough, I could've written
something that uses portage-9999's own internals for this, but i don't
so i didn't. If anyone's bored and wants to re-write/make a new vdb
update script in python, i'd be more than happy to put it on the overlay!
I've opened bug 424429 to help track 4-slot-abi issues as well as
issues with my overlay in general. Please do not file separate bugs.
Ian
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.19 (GNU/Linux)
iF4EAREIAAYFAk/wkGgACgkQ2ugaI38ACPBxGwD8C6oNk8CdQeLsfGJe2bK48UH/
A5aCVbDeBFFxrCjqBqABAJXsf/4sdLKnAj56j3CUukvu5RR5TRXw6U8V0h+a16We
=yIP+
-----END PGP SIGNATURE-----
