| Age | Commit message (Collapse) | Author |
|
- Make ECM_TEST=forceoptional recursive by default
- Leave ECM_TEST=forceoptional-recursive for BUILD_TESTING patch creation
- Unrelated: Drop obsolete extra-cmake-modules condition, it uses cmake.eclass
Signed-off-by: Andreas Sturmlechner <asturm@gentoo.org>
|
|
ECM_PYTHON_BINDINGS was only used since KF6.
Signed-off-by: Andreas Sturmlechner <asturm@gentoo.org>
|
|
Bug: https://bugs.gentoo.org/963981
Fixes: f8c7af1b51019f5101457b35ccdd38d52e4cf1bb
Signed-off-by: Sam James <sam@gentoo.org>
|
|
Define KDEInstallDirs variables to fix ECM >=6.19 warnings.
Use the opportunity and cleanup some quirks including ECM_KDEINSTALLDIRS.
Bug: https://bugs.gentoo.org/963981
Signed-off-by: Andreas Sturmlechner <asturm@gentoo.org>
|
|
Signed-off-by: Andreas Sturmlechner <asturm@gentoo.org>
|
|
Upstream commit 94e3edb259a4cc74545facd49e78867f6ec809fd
Signed-off-by: Andreas Sturmlechner <asturm@gentoo.org>
|
|
Bug: https://bugs.gentoo.org/959142
Signed-off-by: Andreas Sturmlechner <asturm@gentoo.org>
|
|
Upstream now only searches for these Python libs inside the
BUILD_PYTHON_BINDINGS conditional if enabled.
See also:
https://invent.kde.org/frameworks/kunitconversion/-/commit/2c1239dd
Signed-off-by: Andreas Sturmlechner <asturm@gentoo.org>
|
|
6.15 changes how API documentation is built, and we haven't implemented
it yet.
See also:
https://mail.kde.org/pipermail/distributions/2025-June/001595.html
Signed-off-by: Andreas Sturmlechner <asturm@gentoo.org>
|
|
This makes life easier for tinderboxes, but we should improve this in some future
EAPI.
Signed-off-by: Sam James <sam@gentoo.org>
|
|
Signed-off-by: Andreas Sturmlechner <asturm@gentoo.org>
|
|
This omission impressively goes back to initial kde overlay commit
8265e71f48675ee84f8c745734ee157cc74cea6a on 2008-08-04.
Bug: https://bugs.gentoo.org/946863
Signed-off-by: Andreas Sturmlechner <asturm@gentoo.org>
|
|
Allow for whitespace padding within kdoctools_install and ki18n_install.
Odd that this hasn't come up before.
Closes: https://bugs.gentoo.org/950317
Signed-off-by: Andreas Sturmlechner <asturm@gentoo.org>
|
|
Signed-off-by: Andreas Sturmlechner <asturm@gentoo.org>
|
|
Amends c715f7a87d835a5d7eb9204475e4e8360c6e8908
Signed-off-by: Andreas Sturmlechner <asturm@gentoo.org>
|
|
Signed-off-by: Andreas Sturmlechner <asturm@gentoo.org>
|
|
Signed-off-by: Andreas Sturmlechner <asturm@gentoo.org>
|
|
Bug: https://bugs.gentoo.org/813330
Signed-off-by: Andreas Sturmlechner <asturm@gentoo.org>
|
|
Keep inheriting toolchain-funcs.eclass for EAPI-8 only.
Signed-off-by: Andreas Sturmlechner <asturm@gentoo.org>
|
|
Starts with consumers requesting >=KFMIN-6.9.0.
Signed-off-by: Andreas Sturmlechner <asturm@gentoo.org>
|
|
First step towards deprecating this variable.
Signed-off-by: Andreas Sturmlechner <asturm@gentoo.org>
|
|
See also:
https://invent.kde.org/teams/goals/streamlined-application-development-experience/-/issues/9
Signed-off-by: Andreas Sturmlechner <asturm@gentoo.org>
|
|
Other eclass' guard variables are not accessible in global scope.
This (mostly) reverts commit 5f34bb1bd9f2349f3ab3bfd09522b7fbb12c4042.
Signed-off-by: Andreas Sturmlechner <asturm@gentoo.org>
|
|
Not all Frameworks packages are always in kde-frameworks category -
sometimes they are moving - but they all inherit frameworks.kde.org.eclass.
Signed-off-by: Andreas Sturmlechner <asturm@gentoo.org>
|
|
These categories may have sufficed at the creation of ecm.eclass, but
not at this point.
Signed-off-by: Andreas Sturmlechner <asturm@gentoo.org>
|
|
Signed-off-by: Andreas Sturmlechner <asturm@gentoo.org>
|
|
Closes: https://bugs.gentoo.org/823808
Signed-off-by: Andreas Sturmlechner <asturm@gentoo.org>
|
|
Zero current use in ::kde or ::gentoo and highly unlikely anywhere else,
so I think we can get away with it.
Signed-off-by: Andreas Sturmlechner <asturm@gentoo.org>
|
|
Signed-off-by: Andreas Sturmlechner <asturm@gentoo.org>
|
|
Signed-off-by: Andreas Sturmlechner <asturm@gentoo.org>
|
|
Signed-off-by: Andreas Sturmlechner <asturm@gentoo.org>
|
|
Signed-off-by: Andreas Sturmlechner <asturm@gentoo.org>
|
|
Signed-off-by: Andreas Sturmlechner <asturm@gentoo.org>
|
|
- detect if we are inheriting kde.org.eclass
- detect if package name is not the same as upstream name (KDE_ORG_NAME)
- iterate through "${ED}"/usr/share/metainfo/
- if file contains KDE_ORG_NAME or similar, rename that to ${PN}${SLOT/0*/}
- otherwise, prepend "${PN}${SLOT/0*/}-" after ^org.kde.
Notably, this does not (yet?) trigger if a package ${PN} == ${KDE_ORG_NAME}
but different SLOTs. Should we just do that all the time?
Signed-off-by: Andreas Sturmlechner <asturm@gentoo.org>
|
|
Signed-off-by: Andreas Sturmlechner <asturm@gentoo.org>
|
|
Signed-off-by: Andreas Sturmlechner <asturm@gentoo.org>
|
|
But only in KF6-based packages as those are still masked, for getting
away with changing installed files.
Bug: https://bugs.gentoo.org/928345
Signed-off-by: Andreas Sturmlechner <asturm@gentoo.org>
|
|
In dev-qt/qtbase:6, Qt6Test is always provided so it makes no sense to
go through the trouble of ripping out the build dependency.
We should still aim to avoid needless building of tests if not requested,
either by detecting unconditional use of ecm_add_test or expose existing
force* logic to ebuild developers to catch those as errors.
Signed-off-by: Andreas Sturmlechner <asturm@gentoo.org>
|
|
Signed-off-by: Sam James <sam@gentoo.org>
|
|
Signed-off-by: Andreas Sturmlechner <asturm@gentoo.org>
|
|
Signed-off-by: Andreas Sturmlechner <asturm@gentoo.org>
|
|
Signed-off-by: Andreas Sturmlechner <asturm@gentoo.org>
|
|
Signed-off-by: Andreas Sturmlechner <asturm@gentoo.org>
|
|
Signed-off-by: Andreas Sturmlechner <asturm@gentoo.org>
|
|
In ecm.eclass, this only affects BDEPENDs, and in practise every ebuild
making use of KFMIN in RDEPEND will set their own minimum pre-inherit.
Many KDE packages already require 5.101 at a minimum for new API and
porting away from deprecated.
Signed-off-by: Andreas Sturmlechner <asturm@gentoo.org>
|
|
Bug: https://bugs.gentoo.org/836726
Signed-off-by: Andreas Sturmlechner <asturm@gentoo.org>
|
|
Signed-off-by: Andreas Sturmlechner <asturm@gentoo.org>
|
|
Bug: https://bugs.gentoo.org/908715
Signed-off-by: Andreas Sturmlechner <asturm@gentoo.org>
|
|
Keep inheriting virtualx.eclass since it is expected from ecm.eclass for now.
Signed-off-by: Andreas Sturmlechner <asturm@gentoo.org>
|
|
Signed-off-by: Andreas Sturmlechner <asturm@gentoo.org>
|