Re: [gentoo-dev] New eclass: ros-catkin.eclass, new USE_EXPAND: ROS_MESSAGES, new categories: dev-ros and ros-meta for review.

[Alexis Ballier]

On Fri, 18 Sep 2015 10:36:02 +0200
hasufell <hasufell@g.o> wrote:

> How about something like
>
> case ${EAPI} in
> 	5) : ;;
> 	*) die "EAPI=${EAPI:-0} is not supported" ;;
> esac
>
> to force people to stay up2date... then you can also drop most of
> those redunant "|| die"s for EAPI functions
>


ewww yes, thanks, I thought about this but somehow forgot it
afterwards...

just some note: I don't think it's eclasses role to "force" people to
stay up to date, but rather to declare what eapi doesn't work with it,
which in this case is anything before 5.

fixed locally anyway

> >
> > IUSE="test"
> > RDEPEND="
> > 	dev-util/catkin${CATKIN_PYTHON_USEDEP}
> > 	dev-python/empy${CATKIN_PYTHON_USEDEP}
> > "
> > DEPEND="${RDEPEND}"
> >
> > if [ -n "${CATKIN_DO_PYTHON_MULTIBUILD}" ] ; then
> > 	RDEPEND="${RDEPEND} dev-lang/python-exec:2"
> > fi
> >
> > # @ECLASS-VARIABLE: CATKIN_HAS_MESSAGES
> > # @DESCRIPTION:
> > # Set it to a non-empty value before inherit to tell the eclass the
> > package has messages to build. # Messages will be built based on
> > ROS_MESSAGES USE_EXPANDed variable.
> >
> > # @ECLASS-VARIABLE: CATKIN_MESSAGES_TRANSITIVE_DEPS
> > # @DESCRIPTION:
> > # Some messages have dependencies on other messages.
> > # In that case, CATKIN_MESSAGES_TRANSITIVE_DEPS should contain a
> > space-separated list of atoms # representing those dependencies.
> > The eclass uses it to ensure proper dependencies on these packages.
> > if [ -n "${CATKIN_HAS_MESSAGES}" ] ; then IUSE="${IUSE}
> > +ros_messages_python +ros_messages_cxx ros_messages_eus
> > ros_messages_lisp" RDEPEND="${RDEPEND} ros_messages_cxx?
> > ( dev-ros/gencpp:=${CATKIN_PYTHON_USEDEP}  ) ros_messages_eus?
> > ( dev-ros/geneus:=${CATKIN_PYTHON_USEDEP}  ) ros_messages_python?
> > ( dev-ros/genpy:=${CATKIN_PYTHON_USEDEP}   ) ros_messages_lisp?
> > ( dev-ros/genlisp:=${CATKIN_PYTHON_USEDEP} ) dev-ros/message_runtime
> > 	"
> > 	DEPEND="${DEPEND} ${RDEPEND}
> > 		dev-ros/message_generation
> > 		dev-ros/genmsg${CATKIN_PYTHON_USEDEP}
> > 	"
> > 	if [ -n "${CATKIN_MESSAGES_TRANSITIVE_DEPS}" ] ; then
> > 		for i in ${CATKIN_MESSAGES_TRANSITIVE_DEPS} ; do
> > 			ds="${i}[ros_messages_python?,ros_messages_cxx?,ros_messages_lisp?,ros_messages_eus?]
> > ros_messages_python? ( ${i}[${PYTHON_USEDEP}] )"
> > RDEPEND="${RDEPEND} ${ds}" DEPEND="${DEPEND} ${ds}"
> > 		done
> > 	fi
> > fi
> >
>
> Keep in mind what this implies when you change these dependencies
> without bumping the ebuilds that use them.

only way i see these changing is with a new ros_messages_*** useflag,
which will cause a rebuild anyway

> >
> > # @FUNCTION: ros-catkin_src_test_internal
> > # @DESCRIPTION:
> > # Decorator around cmake-utils_src_test to ensure tests are built
> > before running them. ros-catkin_src_test_internal() {
> > 	cd "${BUILD_DIR}"
>
> here is definitely a "|| die" missing, alternatively

yup thx

> emake -C "${BUILD_DIR}" ...
> should work too and it seems cmake-utils_src_test already switches to
> "${BUILD_DIR}" (with missing error handling, lol)
>
> >
> > # @FUNCTION: ros-catkin_src_install_with_python
> > # @DESCRIPTION:
> > # Decorator around cmake-utils_src_install to ensure python scripts
> > are properly handled w.r.t. python-exec2.
> > ros-catkin_src_install_with_python() { python_export
> > PYTHON_SCRIPTDIR cmake-utils_src_install
> > 	if [ ! -f "${T}/.catkin_python_symlinks_generated" -a -d
> > "${D}/${PYTHON_SCRIPTDIR}" ]; then dodir /usr/bin
> > 		for i in "${D}/${PYTHON_SCRIPTDIR}"/* ; do
> > 			dosym ../lib/python-exec/python-exec2
> > "/usr/bin/${i##*/}" || die done
> > 		touch "${T}/.catkin_python_symlinks_generated"
>
> touch can fail too, e.g. if the directory does not exist, so "|| die"
> seems to make sense too

I think $T is guaranteed to exist, but die doesnt hurt, so added
locally too



thx for the review