[Arrow] Another try at building Arrow

A/Arrow/build_tarballs.jl

@@ -0,0 +1,98 @@
+# Note that this script can accept some limited command-line arguments, run
+# `julia build_tarballs.jl --help` to see a usage message.
+using BinaryBuilder, Pkg
+
+name = "Arrow"
+version = v"10.0.0"
+
+# Collection of sources required to complete build
+sources = [
+    ArchiveSource("https://github.com/apache/arrow/archive/refs/tags/apache-arrow-$(version).zip", "42995abe620c41c42b8fbc486f1c63a1e5b1da534718ac66dbc790a88efeaa37")
+    DirectorySource("./bundled")
+]
+
+# Bash recipe for building across all platforms
+script = raw"""
+
+cd $WORKSPACE/srcdir/arrow-apache-arrow-*
+
+# Set toolchain for building external deps
+for f in ${WORKSPACE}/srcdir/patches/*.patch; do
+    atomic_patch -p1 ${f}
+done
+
+cd cpp && mkdir build_dir && cd build_dir
+
+# Ignore check for availibility on older macOS versions
+if [[ "${target}" == x86_64-apple-darwin* ]]; then
+    CXXFLAGS="${CXXFLAGS} -D_LIBCPP_DISABLE_AVAILABILITY"
+fi
+
+CMAKE_FLAGS=(-DCMAKE_INSTALL_PREFIX=$prefix
+-DCMAKE_TOOLCHAIN_FILE=${CMAKE_TARGET_TOOLCHAIN}
+-DARROW_CXXFLAGS="${CXXFLAGS}"
+-DCMAKE_BUILD_TYPE=Release
+-DARROW_BUILD_UTILITIES=OFF
+-DARROW_WITH_UTF8PROC=OFF
+-DARROW_DEPENDENCY_SOURCE=SYSTEM
+-DARROW_VERBOSE_THIRDPARTY_BUILD=ON
+-DARROW_BUILD_STATIC=OFF
+-DARROW_DATASET=ON
+-DARROW_COMPUTE=OFF
+-DARROW_WITH_RE2=OFF
+-DARROW_WITH_BZ2=ON
+-DARROW_IPC=OFF
+-DARROW_WITH_LZ4=ON
+-DARROW_WITH_ZSTD=OFF
+-DARROW_WITH_ZLIB=ON
+-DARROW_WITH_SNAPPY=ON
+-DARROW_THRIFT_USE_SHARED=ON
+-DARROW_PARQUET=ON
+-DPARQUET_BUILD_EXECUTABLES=OFF
+-DARROW_SIMD_LEVEL=NONE
+-DARROW_USE_XSIMD=OFF
+-DARROW_JEMALLOC=OFF
+-Dxsimd_SOURCE=AUTO)
+
+# CMake is doubling the suffixes...
+if [[ "${target}" == *-mingw32 ]]; then
+    ln -s ${prefix}/lib/libthrift.dll.a ${prefix}/lib/libthrift.a.dll.a
+    ln -s ${prefix}/lib/libutf8proc.a ${prefix}/lib/libutf8proc.dll.a.a
+fi
+
+cmake .. "${CMAKE_FLAGS[@]}"
+
+make -j${nproc}
+
+# Remove double suffixes
+if [[ "${target}" == *-mingw32 ]]; then
+    rm ${prefix}/lib/libthrift.a.dll.a
+    rm ${prefix}/lib/libutf8proc.dll.a.a
+fi
+
+make install
+"""
+
+# These are the platforms we will build for by default, unless further
+# platforms are passed in on the command line
+platforms = expand_cxxstring_abis(supported_platforms())
+
+# The products that we will ensure are always built
+products = [
+    LibraryProduct("libparquet", :libparquet),
+    LibraryProduct("libarrow", :libarrow)
+]
+
+# Dependencies that must be installed before this package can be built
+dependencies = [
+    Dependency("boost_jll", compat="=1.76.0")
+    Dependency("Zlib_jll")
+    Dependency("Bzip2_jll", compat="1.0.7")
+    Dependency("Lz4_jll")
+    Dependency("Thrift_jll")
+    Dependency("snappy_jll")
+    Dependency("CompilerSupportLibraries_jll")
+]
+
+# Build the tarballs, and possibly a `build.jl` as well.
+build_tarballs(ARGS, name, version, sources, script, platforms, products, dependencies; julia_compat="1.6", preferred_gcc_version=v"8")

A/Arrow/bundled/patches/cxxflags.patch

@@ -0,0 +1,15 @@
+diff --git a/cpp/cmake_modules/SetupCxxFlags.cmake b/cpp/cmake_modules/SetupCxxFlags.cmake
+index cef4eb0b1..97f866938 100644
+--- a/cpp/cmake_modules/SetupCxxFlags.cmake
++++ b/cpp/cmake_modules/SetupCxxFlags.cmake
+@@ -29,7 +29,9 @@ if(NOT DEFINED ARROW_CPU_FLAG)
+   elseif(CMAKE_SYSTEM_PROCESSOR MATCHES "aarch64|ARM64|arm64")
+     set(ARROW_CPU_FLAG "armv8")
+   elseif(CMAKE_SYSTEM_PROCESSOR MATCHES "armv7")
+-    set(ARROW_CPU_FLAG "armv7")
++    set(ARROW_CPU_FLAG "armv7")
++  elseif(CMAKE_SYSTEM_PROCESSOR MATCHES "armv6")
++    set(ARROW_CPU_FLAG "armv7") # Prevents another patch, will add -latomic to linker
+   elseif(CMAKE_SYSTEM_PROCESSOR MATCHES "powerpc|ppc")
+     set(ARROW_CPU_FLAG "ppc")
+   elseif(CMAKE_SYSTEM_PROCESSOR MATCHES "s390x")

A/Arrow/bundled/patches/thrift.patch

@@ -0,0 +1,12 @@
+diff --git a/cpp/cmake_modules/FindThrift.cmake b/cpp/cmake_modules/FindThrift.cmake
+index 2f20a5cb5..2d1e728aa 100644
+--- a/cpp/cmake_modules/FindThrift.cmake
++++ b/cpp/cmake_modules/FindThrift.cmake
+@@ -146,6 +146,7 @@ if(Thrift_FOUND)
+   endif()
+   set_target_properties(thrift::thrift
+                         PROPERTIES IMPORTED_LOCATION "${THRIFT_LIB}"
++                                   IMPORTED_IMPLIB "${THRIFT_LIB}"
+                                    INTERFACE_INCLUDE_DIRECTORIES "${THRIFT_INCLUDE_DIR}")
+   if(WIN32 AND NOT MSVC_TOOLCHAIN)
+     # We don't need this for Visual C++ because Thrift uses

A/Arrow/bundled/patches/toolchain.patch

@@ -0,0 +1,14 @@
+diff --git a/cpp/cmake_modules/ThirdpartyToolchain.cmake b/cpp/cmake_modules/ThirdpartyToolchain.cmake
+index b7cd31f3d..676f9b93b 100644
+--- a/cpp/cmake_modules/ThirdpartyToolchain.cmake
++++ b/cpp/cmake_modules/ThirdpartyToolchain.cmake
+@@ -753,8 +753,7 @@ endif()
+ # directory. This leads to issues if the variables are exported in a subshell
+ # and the invocation of make/ninja is in distinct subshell without the same
+ # environment (CC/CXX).
+-set(EP_COMMON_TOOLCHAIN -DCMAKE_C_COMPILER=${CMAKE_C_COMPILER}
+-                        -DCMAKE_CXX_COMPILER=${CMAKE_CXX_COMPILER})
++set(EP_COMMON_TOOLCHAIN -DCMAKE_TOOLCHAIN_FILE=$ENV{CMAKE_TARGET_TOOLCHAIN})
+
+ if(CMAKE_AR)
+   set(EP_COMMON_TOOLCHAIN ${EP_COMMON_TOOLCHAIN} -DCMAKE_AR=${CMAKE_AR})