summaryrefslogtreecommitdiff
path: root/eclass
Commit message (Collapse)AuthorAgeFilesLines
* eclass/tests/qmail.sh: eend and return should be separate statementsMike Gilbert2022-04-191-1/+2
| | | | | Closes: https://bugs.gentoo.org/839189 Signed-off-by: Mike Gilbert <floppym@gentoo.org>
* toolchain.eclass: fix S definitionSam James2022-04-191-1/+3
| | | | | | | S isn't _blank_ by default. Fixes: 5a7ea1cef34c3c3026a999d566bc4badc1dccc9c Signed-off-by: Sam James <sam@gentoo.org>
* toolchain.eclass: allow overriding S in ebuildsSam James2022-04-191-10/+12
| | | | | | Useful for RCs-but-we-pretend-they're-not-RCs. Signed-off-by: Sam James <sam@gentoo.org>
* toolchain.eclass: default to .xz for newer patch tarballsSam James2022-04-191-7/+15
| | | | Signed-off-by: Sam James <sam@gentoo.org>
* toolchain.eclass: fix zstd RDEPEND for USE=zstd on gccSam James2022-04-191-1/+2
| | | | Signed-off-by: Sam James <sam@gentoo.org>
* toolchain.eclass: support GCC pre-releases for new branchesSam James2022-04-191-3/+6
| | | | | | Needed for GCC 12. Signed-off-by: Sam James <sam@gentoo.org>
* tree-sitter-grammar.eclass: use edo when linkingMatthew Smith2022-04-191-1/+3
| | | | Signed-off-by: Matthew Smith <matthew@gentoo.org>
* edo.eclass: update eclassdoc to clarify purposeSam James2022-04-181-2/+5
| | | | Signed-off-by: Sam James <sam@gentoo.org>
* edo.eclass: fix eclassdoc typoSam James2022-04-181-1/+1
| | | | Signed-off-by: Sam James <sam@gentoo.org>
* edo.eclass: add new eclassSam James2022-04-181-0/+45
| | | | | Bug: https://bugs.gentoo.org/744880 Signed-off-by: Sam James <sam@gentoo.org>
* qt5-build.eclass: Drop support for Qt 5.15.2Andreas Sturmlechner2022-04-181-51/+9
| | | | Signed-off-by: Andreas Sturmlechner <asturm@gentoo.org>
* eclass/ruby-ng.eclass: add missing eendHans de Graaff2022-04-181-0/+1
| | | | Signed-off-by: Hans de Graaff <graaff@gentoo.org>
* elisp.eclass: Update EAPI conditionalUlrich Müller2022-04-171-2/+2
| | | | Signed-off-by: Ulrich Müller <ulm@gentoo.org>
* elisp-common.eclass: Drop support for EAPI 5Ulrich Müller2022-04-171-14/+5
| | | | | | From Emacs overlay. Signed-off-by: Ulrich Müller <ulm@gentoo.org>
* gnuconfig.eclass: fix eend w/o ebeginSam James2022-04-171-3/+3
| | | | | | | eend should be preceded by an begin call. Closes: https://bugs.gentoo.org/838859 Signed-off-by: Sam James <sam@gentoo.org>
* kernel-2.eclass: fix spellingDavid Seifert2022-04-161-2/+2
| | | | | Closes: https://github.com/gentoo/gentoo/pull/24978 Signed-off-by: David Seifert <soap@gentoo.org>
* kernel-2.eclass: remove EAPI 6David Seifert2022-04-161-14/+12
| | | | Signed-off-by: David Seifert <soap@gentoo.org>
* autotools.eclass: update automake 1.16 stable versionSam James2022-04-161-1/+1
| | | | | Closes: https://bugs.gentoo.org/833654 Signed-off-by: Sam James <sam@gentoo.org>
* java-utils-2.eclass: remove ebegin calls that lack eend callsMike Gilbert2022-04-151-4/+6
| | | | | | | | | | | | Instead, echo the command we are about to run. Closes: https://bugs.gentoo.org/838475 Closes: https://bugs.gentoo.org/838478 Closes: https://bugs.gentoo.org/838481 Closes: https://bugs.gentoo.org/838487 Closes: https://bugs.gentoo.org/838490 Closes: https://bugs.gentoo.org/838493 Signed-off-by: Mike Gilbert <floppym@gentoo.org>
* linux-info.eclass: Call ebegin, properly close with eendThomas Bracht Laumann Jespersen2022-04-151-1/+3
| | | | | | | | | | | | | | | | | A recent QA check added to portage informs of eend being called without a preceding call to ebegin. This warning was emitted during pkg_setup for net-vpn/openvpn, and was traced back to the check_extra_config() function in linux-info.eclass. The preference is here to call ebegin (instead of dropping the lone eend) and in each of the possible exit branches for check_extra_config() call eend appropriately. Reported-by: Sam James <sam@gentoo.org> Signed-off-by: Thomas Bracht Laumann Jespersen <t@laumann.xyz> Closes: https://github.com/gentoo/gentoo/pull/25019 Signed-off-by: Mike Pagano <mpagano@gentoo.org>
* latex-package.eclass: Support EAPI 8Ulrich Müller2022-04-151-18/+23
| | | | | | | | | - BDEPEND instead of DEPEND (partially reverts commit 595611085bc5) - Document TEXMF - Create symlinks with dosym -r - Use debug-print-function Signed-off-by: Ulrich Müller <ulm@gentoo.org>
* latex-package.eclass: Properly quote variablesUlrich Müller2022-04-151-19/+21
| | | | | Bug: https://bugs.gentoo.org/379571 Signed-off-by: Ulrich Müller <ulm@gentoo.org>
* Revert "toolchain.eclass: handle RCs correctly"Sam James2022-04-151-13/+3
| | | | | | | | The install paths are still mixed with 11.2.1/11.3.0. Bleh. This reverts commit 0768e37e87fd7b73566a4f1b658f941b575bca12. Signed-off-by: Sam James <sam@gentoo.org>
* go-module.eclass: fix eclassdoc @DEPRECATION tagsSam James2022-04-151-9/+9
| | | | | Thanks-to: Ionen Wolkens <ionen@gentoo.org> Signed-off-by: Sam James <sam@gentoo.org>
* toolchain.eclass: handle RCs correctlySam James2022-04-151-3/+13
| | | | | | The base GCC-VER file doesn't get updated until the final release. Signed-off-by: Sam James <sam@gentoo.org>
* llvm.org.eclass: add sam to distfile pathsSam James2022-04-131-1/+3
| | | | | | For emergencies. Signed-off-by: Sam James <sam@gentoo.org>
* rebar.eclass: depend on slotted rebar2Matthew Smith2022-04-121-2/+2
| | | | Signed-off-by: Matthew Smith <matthew@gentoo.org>
* autotools.eclass: don't use reserved variable (__)Sam James2022-04-121-4/+4
| | | | | Closes: https://bugs.gentoo.org/806368 Signed-off-by: Sam James <sam@gentoo.org>
* distutils-r1.eclass: don't call copy_egg_info if USE_PEP517=setuptoolsSam James2022-04-111-1/+1
| | | | | | Fixes: 6451cf647f04b31a56ea2dc8c3f84a70c2468f87 Thanks-to: Ionen Wolkens <ionen@gentoo.org> Signed-off-by: Sam James <sam@gentoo.org>
* distutils-r1.eclass: Bump min. dependency versionsMichał Górny2022-04-111-9/+11
| | | | | | | | | Bump min. dependency versions to match oldest stable in ::gentoo, in order to avoid requiring developers to repeat them in BDEPEND with higher version. Closes: https://github.com/gentoo/gentoo/pull/24965 Signed-off-by: Michał Górny <mgorny@gentoo.org>
* distutils-r1.eclass: Print deprecation warning for distutils buildsMichał Górny2022-04-111-0/+9
| | | | | | | | | | Print deprecation warnings for non-PEP517 builds using plain distutils. This is a small subset of Python packages overall, and the first step towards deprecating legacy build support. Transitioning pure distutils packages is also important to avoid .egg-info collisions when we switch from CPython distutils to setuptools-vendored distutils. Signed-off-by: Michał Górny <mgorny@gentoo.org>
* distutils-r1.eclass: Introduce distutils_write_namespaceMichał Górny2022-04-111-0/+41
| | | | | | | | Introduce a distutils_write_namespace helper that can be used to temporarily write a namespace __init__.py as needed to run tests when legacy dev-python/namespace-* packages are installed. Signed-off-by: Michał Górny <mgorny@gentoo.org>
* distutils-r1.eclass: Skip build_ext when there no .c/.pyx filesMichał Górny2022-04-111-1/+11
| | | | | | | | | Skip issuing build_ext when there appears to be no .c/.pyx files. Since starting setuptools is expensive, this gives a major speedup to building pure Python packages. If the check misfires, the worst that can happen is that C extensions will be built serialized. Signed-off-by: Michał Górny <mgorny@gentoo.org>
* distutils-r1.eclass: Call build_ext instead of build in PEP517 modeMichał Górny2022-04-111-1/+5
| | | | | | | | | | | | | | Call `setup.py build_ext` instead of the wholesale `build` in PEP517 mode. After all, the call is not strictly necessary; it is only done in order to build C extensions in parallel (PEP517 backend does not support specifying job count). By skipping the other tasks build command does (and then repeats via PEP517 call), the ebuild can build a few seconds faster. Since this is a potentially breaking change (but very unlikely to actually break anything), let's test it behind GPEP517_TESTING. Signed-off-by: Michał Górny <mgorny@gentoo.org>
* distutils-r1.eclass: Update namespace package doc URLMichał Górny2022-04-111-1/+1
| | | | Signed-off-by: Michał Górny <mgorny@gentoo.org>
* distutils-r1.eclass: Issue a QA warning if "build" existsMichał Górny2022-04-111-1/+7
| | | | Signed-off-by: Michał Górny <mgorny@gentoo.org>
* distutils-r1.eclass: Prefer "rm -rf build" over "setup.py clean -a"Michał Górny2022-04-111-1/+5
| | | | | | | | | | | | | | | | | | Prefer using "rm -rf build" directly over "setup.py clean -a". This has three advantages: 1. It is much faster. 2. It works on packages that have broken "setup.py clean", e.g. dev-python/pydantic. 3. It works on packages that block "setup.py clean" and tell you to use "git clean" (sic!), e.g. dev-python/scipy. This is a potentially (but unlikely) breaking change, so do it conditionally to GPEP517_TESTING. Signed-off-by: Michał Górny <mgorny@gentoo.org>
* eclass/cdrom.eclass: add EAPI 8Conrad Kostecki2022-04-111-3/+3
| | | | Signed-off-by: Conrad Kostecki <conikost@gentoo.org>
* eclass/common-lisp-3: Properly escape recursive install filesMarco Sirabella2022-04-101-1/+3
| | | | | | | | | | | | This is a bit of a roundabout way to do it, but since we want to pass the results of `find` to a bash function, this is what we need to do. All of these functions are a bit convoluted, I believe something similar could be hacked together that looks like `find -type f -name '*.lisp' -exec doins` (or xargs) without the recursive calling or escaping mess Signed-off-by: Marco Sirabella <marco@sirabella.org> Signed-off-by: Ulrich Müller <ulm@gentoo.org>
* cdrom.eclass: fix tabConrad Kostecki2022-04-101-5/+5
| | | | Signed-off-by: Conrad Kostecki <conikost@gentoo.org>
* distutils-r1.eclass: Remove extra deps from jupyter backendMichał Górny2022-04-081-4/+2
| | | | | | | | dev-python/jupyter_packaging depends on these directly, and packages using it don't list these deps in their requires, so I guess it doesn't make sense for us to copy them. Signed-off-by: Michał Górny <mgorny@gentoo.org>
* distutils-r1.eclass: Recognize jupyter PEP517 backendMichał Górny2022-04-081-0/+11
| | | | Signed-off-by: Michał Górny <mgorny@gentoo.org>
* distutils-r1.eclass: Fix PEP517 + single-impl w/ no scriptsMichał Górny2022-04-081-1/+4
| | | | Signed-off-by: Michał Górny <mgorny@gentoo.org>
* mono.eclass: Remove conditionals for EAPIs 0, 1, 2Ulrich Müller2022-04-071-3/+1
| | | | Signed-off-by: Ulrich Müller <ulm@gentoo.org>
* mercurial.eclass: Remove conditionals for EAPIs 0, 1, 2Ulrich Müller2022-04-071-3/+1
| | | | Signed-off-by: Ulrich Müller <ulm@gentoo.org>
* distutils-r1.eclass: Support GPEP517_TESTING modeMichał Górny2022-04-071-31/+57
| | | | | | | | | Support GPEP517_TESTING variable to enable using dev-python/gpep517 instead of inline Python snippets. This is meant to provide the necessary testing before we stabilize it and switch over. Closes: https://github.com/gentoo/gentoo/pull/24910 Signed-off-by: Michał Górny <mgorny@gentoo.org>
* distutils-r1.eclass: Add hatchling support (meh!)Michał Górny2022-04-071-0/+9
| | | | Signed-off-by: Michał Górny <mgorny@gentoo.org>
* eclass/vim-{doc,plugin}.eclass: revert previous commit.Patrice Clement2022-04-062-30/+29
| | | | | | Let's wait until those patches are reviewed on the gentoo-dev ML. Signed-off-by: Patrice Clement <monsieurp@gentoo.org>
* vim-plugin.eclass: support EAPI 8Thomas Bracht Laumann Jespersen2022-04-061-9/+10
| | | | | | | | | | | * Drop EAPI 0, 1, 2 workarounds * Move EXPORT_FUNCTIONS to end of file * Add required @USAGE on functions Bug: https://bugs.gentoo.org/830867 Bug: https://bugs.gentoo.org/830866 Signed-off-by: Thomas Bracht Laumann Jespersen <t@laumann.xyz> Signed-off-by: Patrice Clement <monsieurp@gentoo.org>
* vim-doc.eclass: support EAPI 8Thomas Bracht Laumann Jespersen2022-04-061-20/+20
| | | | | | Signed-off-by: Thomas Bracht Laumann Jespersen <t@laumann.xyz> Closes: https://github.com/gentoo/gentoo/pull/24928 Signed-off-by: Patrice Clement <monsieurp@gentoo.org>