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

["Michał Górny" <mgorny@gentoo.org>]

commit:     5e3b7933ae9613c0d585208dbd8183c61832e194
Author:     Michał Górny <mgorny <AT> gentoo <DOT> org>
AuthorDate: Tue Nov 26 06:10:28 2024 +0000
Commit:     Michał Górny <mgorny <AT> gentoo <DOT> org>
CommitDate: Tue Nov 26 07:01:28 2024 +0000
URL:        https://gitweb.gentoo.org/repo/gentoo.git/commit/?id=5e3b7933

dev-python/pyarrow: Bump to 18.1.0

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

 dev-python/pyarrow/Manifest              |   1 +
 dev-python/pyarrow/pyarrow-18.1.0.ebuild | 112 +++++++++++++++++++++++++++++++
 2 files changed, 113 insertions(+)

diff --git a/dev-python/pyarrow/Manifest b/dev-python/pyarrow/Manifest
index b72958532eff..1df70652cc04 100644
--- a/dev-python/pyarrow/Manifest
+++ b/dev-python/pyarrow/Manifest
@@ -1,3 +1,4 @@
 DIST apache-arrow-18.0.0.tar.gz 19113236 BLAKE2B c358f8e61155432540a4002c6dfba7a3763d934a43eddae0a9a6f98bb26879df51e3cdcc64db6dd684ffde2ef8ef6dcff8906355b5769f71afae907aa1e920c5 SHA512 4df30ab5561da695eaa864422626b9898555d86ca56835c3b8a8ca93a1dbaf081582bb36e2440d1daf7e1dd48c76941f1152a4f25ce0dbcc1c2abe244a00c05e
+DIST apache-arrow-18.1.0.tar.gz 19132388 BLAKE2B 4a8c60d8a5dcf2b8ac4e2ae3a7f555dd0604d5ccc4b72ba317f18e16e75725b24553050fe88a02f382b0581310f20d0b795133b243f2851fa0e8d3801a839358 SHA512 7249c03a6097bc64fb0092143e4d4aaef3227565147e6254f026ddd504177c8dd565a184a0df39743dc989070dc3785e5b66f738c8e310ed9c982b61c2ec4914
 DIST arrow-testing-4d209492d514c2d3cb2d392681b9aa00e6d8da1c.tar.gz 3572558 BLAKE2B fe5b6a46babbde8e2becefea09fecf090201a49faa7d4ee2f44b0bb4da42c344b37564999d32fdbd54a81a56c33fbde9b2d9d2fa664bf127cb5661e845fb6229 SHA512 6758ad936e1eeb2c5dd2b4b40fe90822a48a4ce67ee4d36657c168e1db5d9a17c092f4cf571b13d56b568a773515385441874d1ea36a1f6e5b15a99f67176fa3
 DIST parquet-testing-cb7a9674142c137367bf75a01b79c6e214a73199.tar.gz 1068025 BLAKE2B 0fdd12d04108b27b17564c60ca36e751453f5d74cfbbe3a076e8befa1696aa0f093e2d1a5eacb99e69a98deda502930ba79b8329403af1e8b9d696e6617c6075 SHA512 dd221d9d59442cebb7158d7cf769d6e5323416bfcde2852eee37a26ff089bc04ea10b9ae7158a37d8694c10469d77c7d251d0dd2549596a7b3899d9c39b45997

diff --git a/dev-python/pyarrow/pyarrow-18.1.0.ebuild b/dev-python/pyarrow/pyarrow-18.1.0.ebuild
new file mode 100644
index 000000000000..6203bf592977
--- /dev/null
+++ b/dev-python/pyarrow/pyarrow-18.1.0.ebuild
@@ -0,0 +1,112 @@
+# 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
+
+# arrow.git: testing
+ARROW_DATA_GIT_HASH=4d209492d514c2d3cb2d392681b9aa00e6d8da1c
+# arrow.git: cpp/submodules/parquet-testing
+PARQUET_DATA_GIT_HASH=cb7a9674142c137367bf75a01b79c6e214a73199
+
+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
+	test? (
+		https://github.com/apache/parquet-testing/archive/${PARQUET_DATA_GIT_HASH}.tar.gz
+			-> parquet-testing-${PARQUET_DATA_GIT_HASH}.tar.gz
+		https://github.com/apache/arrow-testing/archive/${ARROW_DATA_GIT_HASH}.tar.gz
+			-> arrow-testing-${ARROW_DATA_GIT_HASH}.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
+		"tests/test_array.py::test_pickling[builtin_pickle]"
+	)
+
+	cd "${T}" || die
+	local -x PYTEST_DISABLE_PLUGIN_AUTOLOAD=1
+	local -x PARQUET_TEST_DATA="${WORKDIR}/parquet-testing-${PARQUET_DATA_GIT_HASH}/data"
+	local -x ARROW_TEST_DATA="${WORKDIR}/arrow-testing-${ARROW_DATA_GIT_HASH}/data"
+	epytest --pyargs pyarrow
+}