{lib}[GCCcore/11.3.0] Arrow v8.0.0, RapidJSON v1.1.0, RE2 v2022-06-01, ... w/ Python 3.10.4

easybuild/easyconfigs/a/Arrow/Arrow-8.0.0-foss-2022.05.eb

@@ -0,0 +1,69 @@
+easyblock = 'CMakeMake'
+
+name = 'Arrow'
+version = '8.0.0'
+
+homepage = 'https://arrow.apache.org'
+description = """Apache Arrow (incl. PyArrow Python bindings), a cross-language development platform
+ for in-memory data."""
+
+toolchain = {'name': 'foss', 'version': '2022.05'}
+
+source_urls = ['https://archive.apache.org/dist/%(namelower)s/%(namelower)s-%(version)s']
+sources = ['apache-arrow-%(version)s.tar.gz']
+checksums = ['ad9a05705117c989c116bae9ac70492fe015050e1b80fb0e38fde4b5d863aaa3']
+
+builddependencies = [
+    ('CMake', '3.23.1'),
+    ('Autotools', '20220317'),
+    ('flex', '2.6.4'),
+    ('Bison', '3.8.2'),
+    ('pkgconf', '1.8.0'),
+]
+
+# Arrow strongly prefers included jemalloc, so not including it as a dependency
+dependencies = [
+    ('Python', '3.10.4'),
+    ('SciPy-bundle', '2022.05'),  # for numpy
+    ('Boost', '1.79.0'),
+    ('lz4', '1.9.3'),
+    ('zlib', '1.2.12'),
+    ('bzip2', '1.0.8'),
+    ('zstd', '1.5.2'),
+    ('snappy', '1.1.9'),
+    ('RapidJSON', '1.1.0'),
+    ('RE2', '2022-06-01'),
+    ('utf8proc', '2.7.0'),
+]
+
+start_dir = 'cpp'
+
+# see https://arrow.apache.org/docs/developers/python.html
+configopts = "-DARROW_PYTHON=on -DARROW_PARQUET=ON -DARROW_WITH_SNAPPY=ON "
+configopts += "-DCMAKE_INSTALL_LIBDIR=lib -DPython3_ROOT_DIR=$EBROOTPYTHON "
+configopts += "-DARROW_WITH_ZLIB=ON -DARROW_WITH_BZ2=ON -DARROW_WITH_ZSTD=ON -DARROW_WITH_LZ4=ON "
+configopts += "-DZSTD_ROOT=$EBROOTZSTD "
+
+# also install Python bindings
+local_install_pyarrow_cmds = "export PKG_CONFIG_PATH=%(installdir)s/lib/pkgconfig:$PKG_CONFIG_PATH && "
+local_install_pyarrow_cmds += "export PYTHONPATH=%(installdir)s/lib/python%(pyshortver)s/site-packages:$PYTHONPATH && "
+local_install_pyarrow_cmds += "cd %(builddir)s/*arrow-%(version)s/python && export XDG_CACHE_HOME=$TMPDIR && "
+local_install_pyarrow_cmds += "sed -i 's/numpy==[0-9.]*/numpy/g' pyproject.toml && "
+local_install_pyarrow_cmds += "Python3_ROOT_DIR=$EBROOTPYTHON "
+local_install_pyarrow_cmds += "PYARROW_WITH_PARQUET=1 pip install --prefix %(installdir)s ."
+postinstallcmds = [local_install_pyarrow_cmds]
+
+modextrapaths = {'PYTHONPATH': 'lib/python%(pyshortver)s/site-packages'}
+
+sanity_check_paths = {
+    'files': ['lib/libarrow.a', 'lib/libarrow.%s' % SHLIB_EXT,
+              'lib/libarrow_python.a', 'lib/libarrow_python.%s' % SHLIB_EXT],
+    'dirs': ['include/arrow', 'lib/cmake/arrow', 'lib/pkgconfig', 'lib/python%(pyshortver)s/site-packages'],
+}
+
+sanity_check_commands = [
+    "python -c 'import pyarrow'",
+    "python -c 'import pyarrow.parquet'",
+]
+
+moduleclass = 'data'

easybuild/easyconfigs/r/RE2/RE2-2022-06-01-GCCcore-11.3.0.eb

@@ -0,0 +1,33 @@
+##
+# Author:    Robert Mijakovic <[email protected]>
+##
+easyblock = "CMakeMake"
+
+name = 'RE2'
+version = '2022-06-01'
+
+homepage = 'https://github.com/google/re2'
+description = """
+RE2 is a fast, safe, thread-friendly alternative to backtracking regular
+expression engines like those used in PCRE, Perl, and Python. It is a C++
+library. """
+
+toolchain = {'name': 'GCCcore', 'version': '11.3.0'}
+toolchainopts = {'pic': True}
+
+github_account = 'google'
+source_urls = [GITHUB_SOURCE]
+sources = ['%(version)s.tar.gz']
+checksums = ['f89c61410a072e5cbcf8c27e3a778da7d6fd2f2b5b1445cd4f4508bee946ab0f']
+
+builddependencies = {
+    ('binutils', '2.38'),
+    ('CMake', '3.23.1'),
+}
+
+sanity_check_paths = {
+    'files': ['lib/libre2.a'],
+    'dirs': ['include/re2'],
+}
+
+moduleclass = 'lib'

easybuild/easyconfigs/r/RapidJSON/RapidJSON-1.1.0-GCCcore-11.3.0.eb

@@ -0,0 +1,30 @@
+easyblock = 'CMakeMake'
+
+name = 'RapidJSON'
+version = '1.1.0'
+
+homepage = 'https://rapidjson.org'
+description = "A fast JSON parser/generator for C++ with both SAX/DOM style API"
+
+toolchain = {'name': 'GCCcore', 'version': '11.3.0'}
+
+source_urls = ['https://github.com/Tencent/%(namelower)s/archive/']
+sources = ['v%(version)s.tar.gz']
+patches = ['%(name)s-%(version)s_fix-compiler-errors.patch']
+checksums = [
+    'bf7ced29704a1e696fbccf2a2b4ea068e7774fa37f6d7dd4039d0787f8bed98e',  # v1.1.0.tar.gz
+    # %(name)s-%(version_major_minor)s.0_fix-compiler-errors.patch
+    '2e40ef6c46bf355feac16cd84faad5b7aca0ad54273117b85534b1b781f411c7',
+]
+
+builddependencies = [
+    ('binutils', '2.38'),
+    ('CMake', '3.23.1'),
+]
+
+sanity_check_paths = {
+    'files': ['lib/pkgconfig/%(name)s.pc'],
+    'dirs': ['include/%(namelower)s', 'lib/cmake', 'share'],
+}
+
+moduleclass = 'lib'

easybuild/easyconfigs/s/snappy/snappy-1.1.9-GCCcore-11.3.0.eb

@@ -0,0 +1,39 @@
+easyblock = 'CMakeMake'
+
+name = 'snappy'
+version = '1.1.9'
+
+homepage = 'https://github.com/google/snappy'
+description = """Snappy is a compression/decompression library. It does not aim
+for maximum compression, or compatibility with any other compression library;
+instead, it aims for very high speeds and reasonable compression."""
+
+toolchain = {'name': 'GCCcore', 'version': '11.3.0'}
+
+source_urls = ['https://github.com/google/snappy/archive/']
+sources = ['%(version)s.tar.gz']
+patches = [
+    '%(name)s-%(version)s_inline-functions.patch',
+    '%(name)s-%(version)s_use-default-rtti.patch',
+]
+checksums = [
+    '75c1fbb3d618dd3a0483bff0e26d0a92b495bbe5059c8b4f1c962b478b6e06e7',  # 1.1.9.tar.gz
+    'ad79190b274df5ddabf14eddd2bb0d9a091ee7d44e4afde89febf9a8f783fdce',  # snappy-1.1.9_inline-functions.patch
+    'af56538330b2d781677c7d94576c15fc36e004ae0b4f1ac7d86bbec22b65e73d',  # snappy-1.1.9_use-default-rtti.patch
+]
+
+builddependencies = [
+    ('binutils', '2.38'),
+    ('CMake', '3.23.1'),
+]
+
+# Disable building tests and benchmarks - we're not using them and they require googletest and benchmark source code
+_configopts = '-DSNAPPY_BUILD_TESTS=OFF -DSNAPPY_BUILD_BENCHMARKS=OFF'
+configopts = ['%s' % _configopts, '-DBUILD_SHARED_LIBS=ON %s' % _configopts]
+
+sanity_check_paths = {
+    'files': ['lib64/libsnappy.a', 'lib64/libsnappy.%s' % SHLIB_EXT, 'include/snappy.h'],
+    'dirs': []
+}
+
+moduleclass = 'lib'

easybuild/easyconfigs/u/utf8proc/utf8proc-2.7.0-GCCcore-11.3.0.eb

@@ -0,0 +1,30 @@
+easyblock = 'CMakeMake'
+
+name = 'utf8proc'
+version = '2.7.0'
+
+homepage = 'https://github.com/JuliaStrings/utf8proc'
+description = """utf8proc is a small, clean C library that provides Unicode normalization, case-folding, 
+and other operations for data in the UTF-8 encoding."""
+
+toolchain = {'name': 'GCCcore', 'version': '11.3.0'}
+
+source_urls = ['https://github.com/JuliaStrings/utf8proc/archive/']
+sources = ['v%(version)s.tar.gz']
+checksums = ['4bb121e297293c0fd55f08f83afab6d35d48f0af4ecc07523ad8ec99aa2b12a1']
+
+builddependencies = [
+    ('binutils', '2.38'),
+    ('CMake', '3.23.1'),
+]
+
+separate_build_dir = True
+
+configopts = ['', '-DBUILD_SHARED_LIBS=true']
+
+sanity_check_paths = {
+    'files': ['include/utf8proc.h', 'lib/libutf8proc.a', 'lib/libutf8proc.%s' % SHLIB_EXT],
+    'dirs': [],
+}
+
+moduleclass = 'lib'