[gentoo-commits] repo/gentoo:master commit in: dev-python/pyarrow/

["Michał Górny" <mgorny@gentoo.org>]

commit:     e28a751cecf48861af85e6829b2bd845fc86401a
Author:     Michał Górny <mgorny <AT> gentoo <DOT> org>
AuthorDate: Wed Jul 17 11:02:16 2024 +0000
Commit:     Michał Górny <mgorny <AT> gentoo <DOT> org>
CommitDate: Wed Jul 17 12:35:16 2024 +0000
URL:        https://gitweb.gentoo.org/repo/gentoo.git/commit/?id=e28a751c

dev-python/pyarrow: Bump to 17.0.0

Signed-off-by: Michał Górny <mgorny <AT> gentoo.org>

 dev-python/pyarrow/Manifest              |   1 +
 dev-python/pyarrow/pyarrow-17.0.0.ebuild | 102 +++++++++++++++++++++++++++++++
 2 files changed, 103 insertions(+)

diff --git a/dev-python/pyarrow/Manifest b/dev-python/pyarrow/Manifest
index b48d5052258a..b4f800fff015 100644
--- a/dev-python/pyarrow/Manifest
+++ b/dev-python/pyarrow/Manifest
@@ -1 +1,2 @@
 DIST apache-arrow-16.1.0.tar.gz 21707079 BLAKE2B 5541c364351e59e8164f2461546bce3ddc834d70b3a03b815b681187674bdd89ccfb7eb113bf9010c821845b0c1fa51f5321bfea5b20927e6364f6452ec980d2 SHA512 28975f59e1fdde2dba4afaf4a5ba934b63db3a7f27656e2aa0af0f0d2a046c9dbfa9a6082de94629c36d03809b296566a37ea65ec5a2fc17fedac7d21e272d31
+DIST apache-arrow-17.0.0.tar.gz 21822331 BLAKE2B 64f9321cbf13fafae5938f26840d3ef642d6f5b40386315030bc70fe8703786ee394c8483d2bde7961cf2b5c15470ffebd7c66029e116b05cfbf5524e9fe51f1 SHA512 4e2a617b8deeb9f94ee085653a721904a75696f0827bcba82b535cc7f4f723066a09914c7fa83c593e51a8a4031e8bf99e563cac1ebb1d89604cb406975d4864

diff --git a/dev-python/pyarrow/pyarrow-17.0.0.ebuild b/dev-python/pyarrow/pyarrow-17.0.0.ebuild
new file mode 100644
index 000000000000..1c2d5fde59a7
--- /dev/null
+++ b/dev-python/pyarrow/pyarrow-17.0.0.ebuild
@@ -0,0 +1,102 @@
+# Copyright 2023-2024 Gentoo Authors
+# Distributed under the terms of the GNU General Public License v2
+
+EAPI=8
+
+DISTUTILS_EXT=1
+DISTUTILS_USE_PEP517=setuptools
+PYTHON_COMPAT=( python3_{10..13} )
+
+inherit distutils-r1 multiprocessing
+
+DESCRIPTION="Python library for Apache Arrow"
+HOMEPAGE="
+	https://arrow.apache.org/
+	https://github.com/apache/arrow/
+	https://pypi.org/project/pyarrow/
+"
+SRC_URI="mirror://apache/arrow/arrow-${PV}/apache-arrow-${PV}.tar.gz"
+S="${WORKDIR}/apache-arrow-${PV}/python"
+
+LICENSE="Apache-2.0"
+SLOT="0"
+KEYWORDS="~amd64 ~arm64 ~loong ~riscv ~x86"
+IUSE="+parquet +snappy ssl"
+
+RDEPEND="
+	~dev-libs/apache-arrow-${PV}[compute,dataset,json,parquet?,re2,snappy?,ssl?]
+	>=dev-python/numpy-1.16.6:=[${PYTHON_USEDEP}]
+"
+BDEPEND="
+	test? (
+		dev-python/cffi[${PYTHON_USEDEP}]
+		dev-python/hypothesis[${PYTHON_USEDEP}]
+		dev-python/pandas[${PYTHON_USEDEP}]
+		dev-python/pytz[${PYTHON_USEDEP}]
+		dev-libs/apache-arrow[lz4,zlib]
+	)
+"
+
+EPYTEST_XDIST=1
+distutils_enable_tests pytest
+
+src_prepare() {
+	distutils-r1_src_prepare
+
+	# cython's -Werror
+	sed -i -e '/--warning-errors/d' CMakeLists.txt || die
+}
+
+src_compile() {
+	export PYARROW_PARALLEL="$(makeopts_jobs)"
+	export PYARROW_BUILD_VERBOSE=1
+	export PYARROW_CXXFLAGS="${CXXFLAGS}"
+	export PYARROW_BUNDLE_ARROW_CPP_HEADERS=0
+	export PYARROW_CMAKE_GENERATOR=Ninja
+	export PYARROW_WITH_HDFS=1
+	if use parquet; then
+		export PYARROW_WITH_DATASET=1
+		export PYARROW_WITH_PARQUET=1
+		use ssl && export PYARROW_WITH_PARQUET_ENCRYPTION=1
+	fi
+	if use snappy; then
+		export PYARROW_WITH_SNAPPY=1
+	fi
+
+	distutils-r1_src_compile
+}
+
+python_test() {
+	local EPYTEST_DESELECT=(
+		# wtf?
+		tests/test_fs.py::test_localfs_errors
+		# these require apache-arrow with jemalloc that doesn't seem
+		# to be supported by the Gentoo package
+		tests/test_memory.py::test_env_var
+		tests/test_memory.py::test_specific_memory_pools
+		tests/test_memory.py::test_supported_memory_backends
+		# hypothesis health check failures
+		# https://github.com/apache/arrow/issues/41318
+		tests/interchange/test_interchange_spec.py::test_dtypes
+		tests/test_convert_builtin.py::test_array_to_pylist_roundtrip
+		tests/test_feather.py::test_roundtrip
+		tests/test_pandas.py::test_array_to_pandas_roundtrip
+		tests/test_strategies.py::test_types
+		tests/test_types.py::test_hashing
+		# fragile memory tests
+		tests/test_csv.py::TestSerialStreamingCSVRead::test_batch_lifetime
+		tests/test_csv.py::TestThreadedStreamingCSVRead::test_batch_lifetime
+		# takes forever, and manages to generate timedeltas over 64 bits
+		tests/test_strategies.py
+	)
+
+	cd "${T}" || die
+	local -x PYTEST_DISABLE_PLUGIN_AUTOLOAD=1
+	epytest --pyargs pyarrow
+}
+
+python_install() {
+	distutils-r1_python_install
+	# https://github.com/apache/arrow/issues/43299
+	rm -r "${D}$(python_get_sitedir)"/{benchmarks,cmake_modules,examples,scripts} || die
+}