apache · coryan · Sep 29, 2021 · Sep 30, 2021 · Oct 2, 2021 · Oct 4, 2021
diff --git a/ci/docker/ubuntu-20.04-cpp.dockerfile b/ci/docker/ubuntu-20.04-cpp.dockerfile
@@ -110,6 +110,7 @@ ENV ARROW_BUILD_TESTS=ON \
     ARROW_DATASET=ON \
     ARROW_FLIGHT=OFF \
     ARROW_GANDIVA=ON \
+    ARROW_GCS=ON \
     ARROW_HDFS=ON \
     ARROW_HOME=/usr/local \
     ARROW_INSTALL_NAME_RPATH=OFF \
@@ -129,6 +130,7 @@ ENV ARROW_BUILD_TESTS=ON \
     ARROW_WITH_ZSTD=ON \
     ASAN_SYMBOLIZER_PATH=/usr/lib/llvm-${llvm}/bin/llvm-symbolizer \
     AWSSDK_SOURCE=BUNDLED \
+    google_cloud_cpp_storage_SOURCE=BUNDLED \
     GTest_SOURCE=BUNDLED \
     gRPC_SOURCE=BUNDLED \
     ORC_SOURCE=BUNDLED \

diff --git a/ci/scripts/cpp_build.sh b/ci/scripts/cpp_build.sh
@@ -74,6 +74,7 @@ cmake -G "${CMAKE_GENERATOR:-Ninja}" \
       -DARROW_GANDIVA_JAVA=${ARROW_GANDIVA_JAVA:-OFF} \
       -DARROW_GANDIVA_PC_CXX_FLAGS=${ARROW_GANDIVA_PC_CXX_FLAGS:-} \
       -DARROW_GANDIVA=${ARROW_GANDIVA:-OFF} \
+      -DARROW_GCS=${ARROW_GCS:-OFF} \
       -DARROW_HDFS=${ARROW_HDFS:-ON} \
       -DARROW_HIVESERVER2=${ARROW_HIVESERVER2:-OFF} \
       -DARROW_INSTALL_NAME_RPATH=${ARROW_INSTALL_NAME_RPATH:-ON} \
@@ -121,6 +122,7 @@ cmake -G "${CMAKE_GENERATOR:-Ninja}" \
       -DCMAKE_INSTALL_PREFIX=${CMAKE_INSTALL_PREFIX:-${ARROW_HOME}} \
       -DCMAKE_UNITY_BUILD=${CMAKE_UNITY_BUILD:-OFF} \
       -Dgflags_SOURCE=${gflags_SOURCE:-} \
+      -Dgoogle_cloud_cpp_storage_SOURCE=${google_cloud_cpp_storage_SOURCE:-} \
       -DgRPC_SOURCE=${gRPC_SOURCE:-} \
       -DGTest_SOURCE=${GTest_SOURCE:-} \
       -DLz4_SOURCE=${Lz4_SOURCE:-} \

diff --git a/cpp/CMakeLists.txt b/cpp/CMakeLists.txt
@@ -740,6 +740,14 @@ if(ARROW_ORC)
   endif()
 endif()
 
+if(ARROW_GCS)
+  list(APPEND ARROW_LINK_LIBS google-cloud-cpp::storage)
+  list(APPEND ARROW_STATIC_LINK_LIBS google-cloud-cpp::storage)
+  if(google_cloud_cpp_storage_SOURCE STREQUAL "SYSTEM")
+    list(APPEND ARROW_STATIC_INSTALL_INTERFACE_LIBS google-cloud-cpp::storage)
+  endif()
+endif()
+
 if(ARROW_USE_GLOG)
   list(APPEND ARROW_LINK_LIBS glog::glog)
   list(APPEND ARROW_STATIC_LINK_LIBS glog::glog)

diff --git a/cpp/cmake_modules/DefineOptions.cmake b/cpp/cmake_modules/DefineOptions.cmake
@@ -224,6 +224,10 @@ if("${CMAKE_SOURCE_DIR}" STREQUAL "${CMAKE_CURRENT_SOURCE_DIR}")
 
   define_option(ARROW_GANDIVA "Build the Gandiva libraries" OFF)
 
+  define_option(ARROW_GCS
+                "Build Arrow with GCS support (requires the GCloud SDK for C++)" OFF)
+  mark_as_advanced(ARROW_GCS) # TODO(ARROW-1231) - remove once completed
+
   define_option(ARROW_HDFS "Build the Arrow HDFS bridge" OFF)
 
   define_option(ARROW_HIVESERVER2 "Build the HiveServer2 client and Arrow adapter" OFF)

diff --git a/cpp/cmake_modules/ThirdpartyToolchain.cmake b/cpp/cmake_modules/ThirdpartyToolchain.cmake
@@ -151,6 +151,8 @@ macro(build_dependency DEPENDENCY_NAME)
     build_gflags()
   elseif("${DEPENDENCY_NAME}" STREQUAL "GLOG")
     build_glog()
+  elseif("${DEPENDENCY_NAME}" STREQUAL "google_cloud_cpp_storage")
+    build_google_cloud_cpp_storage()
   elseif("${DEPENDENCY_NAME}" STREQUAL "gRPC")
     build_grpc()
   elseif("${DEPENDENCY_NAME}" STREQUAL "GTest")
@@ -287,6 +289,10 @@ if(ARROW_FLIGHT)
   set(ARROW_WITH_ZLIB ON)
 endif()
 
+if(ARROW_GCS)
+  set(ARROW_WITH_GOOGLE_CLOUD_CPP ON)
+endif()
+
 if(ARROW_JSON)
   set(ARROW_WITH_RAPIDJSON ON)
 endif()
@@ -433,6 +439,14 @@ else()
   )
 endif()
 
+if(DEFINED ENV{ARROW_CRC32C_URL})
+  set(CRC32C_URL "$ENV{ARROW_CRC32C_URL}")
+else()
+  set_urls(CRC32C_SOURCE_URL
+           "https://github.com/google/crc32c/archive/${ARROW_CRC32C_BUILD_VERSION}.tar.gz"
+  )
+endif()
+
 if(DEFINED ENV{ARROW_GBENCHMARK_URL})
   set(GBENCHMARK_SOURCE_URL "$ENV{ARROW_GBENCHMARK_URL}")
 else()
@@ -460,6 +474,14 @@ else()
   )
 endif()
 
+if(DEFINED ENV{ARROW_GOOGLE_CLOUD_CPP_URL})
+  set(google_cloud_cpp_storage_SOURCE_URL "$ENV{ARROW_GOOGLE_CLOUD_CPP_URL}")
+else()
+  set_urls(google_cloud_cpp_storage_SOURCE_URL
+           "https://github.com/googleapis/google-cloud-cpp/archive/${ARROW_GOOGLE_CLOUD_CPP_BUILD_VERSION}.tar.gz"
+  )
+endif()
+
 if(DEFINED ENV{ARROW_GRPC_URL})
   set(GRPC_SOURCE_URL "$ENV{ARROW_GRPC_URL}")
 else()
@@ -497,6 +519,14 @@ else()
   )
 endif()
 
+if(DEFINED ENV{ARROW_NLOHMANN_JSON_URL})
+  set(NLOHMANN_JSON_SOURCE_URL "$ENV{ARROW_NLOHMANN_JSON_URL}")
+else()
+  set_urls(NLOHMANN_JSON_SOURCE_URL
+           "https://github.com/nlohmann/json/archive/${ARROW_NLOHMANN_JSON_BUILD_VERSION}.tar.gz"
+  )
+endif()
+
 if(DEFINED ENV{ARROW_LZ4_URL})
   set(LZ4_SOURCE_URL "$ENV{ARROW_LZ4_URL}")
 else()
@@ -3470,6 +3500,197 @@ if(ARROW_WITH_GRPC)
   endif()
 endif()
 
+# ----------------------------------------------------------------------
+# GCS and dependencies
+
+macro(build_crc32c_once)
+  if(NOT TARGET crc32c_ep)
+    message(STATUS "Building crc32c from source")
+    # Build crc32c
+    set(CRC32C_PREFIX "${CMAKE_CURRENT_BINARY_DIR}/crc32c_ep-install")
+    set(CRC32C_CMAKE_ARGS
+        ${EP_COMMON_CMAKE_ARGS}
+        -DCMAKE_INSTALL_LIBDIR=lib
+        "-DCMAKE_INSTALL_PREFIX=<INSTALL_DIR>"
+        -DCMAKE_CXX_STANDARD=11
+        -DCRC32C_BUILD_TESTS=OFF
+        -DCRC32C_BUILD_BENCHMARKS=OFF
+        -DCRC32C_USE_GLOG=OFF)
+
+    set(_CRC32C_STATIC_LIBRARY
+        "${CRC32C_PREFIX}/lib/${CMAKE_STATIC_LIBRARY_PREFIX}crc32c${CMAKE_STATIC_LIBRARY_SUFFIX}"
+    )
+    set(CRC32C_BUILD_BYPRODUCTS ${_CRC32C_STATIC_LIBRARY})
+    set(CRC32C_LIBRARIES crc32c)
+
+    externalproject_add(crc32c_ep
+                        ${EP_LOG_OPTIONS}
+                        INSTALL_DIR ${CRC32C_PREFIX}
+                        URL ${CRC32C_SOURCE_URL}
+                        URL_HASH "SHA256=${ARROW_CRC32C_BUILD_SHA256_CHECKSUM}"
+                        CMAKE_ARGS ${CRC32C_CMAKE_ARGS}
+                        BUILD_BYPRODUCTS ${CRC32C_BUILD_BYPRODUCTS})
+    add_library(Crc32c::crc32c STATIC IMPORTED)
+    set_target_properties(Crc32c::crc32c
+                          PROPERTIES IMPORTED_LOCATION ${_CRC32C_STATIC_LIBRARY}
+                                     INTERFACE_INCLUDE_DIRECTORIES
+                                     "${CRC32C_INCLUDE}/include")
+    add_dependencies(Crc32c::crc32c crc32c_ep)
+  endif()
+endmacro()
+
+macro(build_nlohmann_json_once)
+  if(NOT TARGET nlohmann_json_ep)
+    message(STATUS "Building nlohmann-json from source")
+    # "Build" nlohmann-json
+    set(NLOHMANN_JSON_PREFIX "${CMAKE_CURRENT_BINARY_DIR}/nlohmann_json_ep-install")
+    set(NLOHMANN_JSON_CMAKE_ARGS
+        ${EP_COMMON_CMAKE_ARGS} -DCMAKE_CXX_STANDARD=11
+        "-DCMAKE_INSTALL_PREFIX=<INSTALL_DIR>" -DBUILD_TESTING=OFF)
+
+    set(NLOHMANN_JSON_BUILD_BYPRODUCTS ${NLOHMANN_JSON_PREFIX}/include/json.hpp)
+
+    externalproject_add(nlohmann_json_ep
+                        ${EP_LOG_OPTIONS}
+                        INSTALL_DIR ${NLOHMANN_JSON_PREFIX}
+                        URL ${NLOHMANN_JSON_SOURCE_URL}
+                        URL_HASH "SHA256=${ARROW_NLOHMANN_JSON_BUILD_SHA256_CHECKSUM}"
+                        CMAKE_ARGS ${NLOHMANN_JSON_CMAKE_ARGS}
+                        BUILD_BYPRODUCTS ${NLOHMANN_JSON_BUILD_BYPRODUCTS})
+    add_library(nlohmann_json::nlohmann_json INTERFACE IMPORTED)
+    set_target_properties(nlohmann_json::nlohmann_json
+                          PROPERTIES INTERFACE_INCLUDE_DIRECTORIES
+                                     "${NLOHMANN_JSON_PREFIX}/include")
+    add_dependencies(nlohmann_json::nlohmann_json nlohmann_json_ep)
+  endif()
+endmacro()
+
+macro(build_google_cloud_cpp_storage)
+  message(STATUS "Building google-cloud-cpp from source")
+  message(STATUS "Only building the google-cloud-cpp::storage component")
+
+  # List of dependencies taken from https://github.com/googleapis/google-cloud-cpp/blob/master/doc/packaging.md
+  build_absl_once()
+  build_crc32c_once()
+  build_nlohmann_json_once()
+
+  # Curl is required on all platforms, but building it internally might also trip over S3's copy.
+  # For now, force its inclusion from the underlying system or fail.
+  find_package(CURL REQUIRED 7.47.0)
+
+  # Build google-cloud-cpp, with only storage_client
+
+  # Inject vendored packages via CMAKE_PREFIX_PATH
+  list(APPEND GOOGLE_CLOUD_CPP_PREFIX_PATH_LIST ${ABSL_PREFIX})
+  list(APPEND GOOGLE_CLOUD_CPP_PREFIX_PATH_LIST ${CRC32C_PREFIX})
+  list(APPEND GOOGLE_CLOUD_CPP_PREFIX_PATH_LIST ${NLOHMANN_JSON_PREFIX})
+
+  set(GOOGLE_CLOUD_CPP_PREFIX_PATH_LIST_SEP_CHAR "|")
+  list(JOIN GOOGLE_CLOUD_CPP_PREFIX_PATH_LIST
+       ${GOOGLE_CLOUD_CPP_PREFIX_PATH_LIST_SEP_CHAR} GOOGLE_CLOUD_CPP_PREFIX_PATH)
+
+  set(GOOGLE_CLOUD_CPP_INSTALL_PREFIX
+      "${CMAKE_CURRENT_BINARY_DIR}/google_cloud_cpp_ep-install")
+  set(GOOGLE_CLOUD_CPP_INCLUDE_DIR "${GOOGLE_CLOUD_CPP_INSTALL_PREFIX}/include")
+  set(GOOGLE_CLOUD_CPP_CMAKE_ARGS
+      ${EP_COMMON_CMAKE_ARGS}
+      -DBUILD_TESTING=OFF
+      -DCMAKE_INSTALL_LIBDIR=lib
+      "-DCMAKE_INSTALL_PREFIX=<INSTALL_DIR>"
+      -DCMAKE_INSTALL_RPATH=$ORIGIN
+      -DCMAKE_PREFIX_PATH=${GOOGLE_CLOUD_CPP_PREFIX_PATH}
+      # Compile only the storage library and its dependencies. To enable
+      # other services (Spanner, Bigtable, etc.) add them (as a list) to this
+      # parameter. Each has its own `google-cloud-cpp::*` library.
+      -DGOOGLE_CLOUD_CPP_ENABLE=storage)
+  if(OPENSSL_ROOT_DIR)
+    list(APPEND GOOGLE_CLOUD_CPP_CMAKE_ARGS -DOPENSSL_ROOT_DIR=${OPENSSL_ROOT_DIR})
+  endif()
+
+  add_custom_target(google_cloud_cpp_dependencies)
+
+  add_dependencies(google_cloud_cpp_dependencies absl_ep)
+  add_dependencies(google_cloud_cpp_dependencies crc32c_ep)
+  add_dependencies(google_cloud_cpp_dependencies nlohmann_json_ep)
+  # Typically the steps to build the AWKSSDK provide `CURL::libcurl`, but if that is
+  # disabled we need to provide our own.
+  if(NOT TARGET CURL::libcurl)
+    find_package(CURL REQUIRED)
+    if(NOT TARGET CURL::libcurl)
+      # For CMake 3.11 or older
+      add_library(CURL::libcurl UNKNOWN IMPORTED)
+      set_target_properties(CURL::libcurl
+                            PROPERTIES INTERFACE_INCLUDE_DIRECTORIES
+                                       "${CURL_INCLUDE_DIRS}" IMPORTED_LOCATION
+                                                              "${CURL_LIBRARIES}")
+    endif()
+  endif()
+
+  set(GOOGLE_CLOUD_CPP_STATIC_LIBRARY_STORAGE
+      "${GOOGLE_CLOUD_CPP_INSTALL_PREFIX}/lib/${CMAKE_STATIC_LIBRARY_PREFIX}google_cloud_cpp_storage${CMAKE_STATIC_LIBRARY_SUFFIX}"
+  )
+
+  set(GOOGLE_CLOUD_CPP_STATIC_LIBRARY_COMMON
+      "${GOOGLE_CLOUD_CPP_INSTALL_PREFIX}/lib/${CMAKE_STATIC_LIBRARY_PREFIX}google_cloud_cpp_common${CMAKE_STATIC_LIBRARY_SUFFIX}"
+  )
+
+  externalproject_add(google_cloud_cpp_ep
+                      ${EP_LOG_OPTIONS}
+                      LIST_SEPARATOR ${GOOGLE_CLOUD_CPP_PREFIX_PATH_LIST_SEP_CHAR}
+                      INSTALL_DIR ${GOOGLE_CLOUD_CPP_INSTALL_PREFIX}
+                      URL ${google_cloud_cpp_storage_SOURCE_URL}
+                      URL_HASH "SHA256=${ARROW_GOOGLE_CLOUD_CPP_BUILD_SHA256_CHECKSUM}"
+                      CMAKE_ARGS ${GOOGLE_CLOUD_CPP_CMAKE_ARGS}
+                      BUILD_BYPRODUCTS ${GOOGLE_CLOUD_CPP_STATIC_LIBRARY_STORAGE}
+                                       ${GOOGLE_CLOUD_CPP_STATIC_LIBRARY_COMMON}
+                      DEPENDS google_cloud_cpp_dependencies)
+  add_dependencies(toolchain google_cloud_cpp_ep)
+
+  add_library(google-cloud-cpp::common STATIC IMPORTED)
+  set_target_properties(google-cloud-cpp::common
+                        PROPERTIES IMPORTED_LOCATION
+                                   "${GOOGLE_CLOUD_CPP_STATIC_LIBRARY_COMMON}"
+                                   INTERFACE_INCLUDE_DIRECTORIES
+                                   "${GOOGLE_CLOUD_CPP_INCLUDE_DIR}")
+  set_property(TARGET google-cloud-cpp::common
+               PROPERTY INTERFACE_LINK_LIBRARIES
+                        absl::any
+                        absl::flat_hash_map
+                        absl::memory
+                        absl::optional
+                        absl::time
+                        Threads::Threads)
+
+  add_library(google-cloud-cpp::storage STATIC IMPORTED)
+  set_target_properties(google-cloud-cpp::storage
+                        PROPERTIES IMPORTED_LOCATION
+                                   "${GOOGLE_CLOUD_CPP_STATIC_LIBRARY_STORAGE}"
+                                   INTERFACE_INCLUDE_DIRECTORIES
+                                   "${GOOGLE_CLOUD_CPP_INCLUDE_DIR}")
+  set_property(TARGET google-cloud-cpp::storage
+               PROPERTY INTERFACE_LINK_LIBRARIES
+                        google-cloud-cpp::common
+                        absl::memory
+                        absl::strings
+                        absl::str_format
+                        absl::time
+                        absl::variant
+                        nlohmann_json::nlohmann_json
+                        Crc32c::crc32c
+                        CURL::libcurl
+                        Threads::Threads
+                        OpenSSL::SSL
+                        OpenSSL::Crypto)
+  add_dependencies(google-cloud-cpp::storage google_cloud_cpp_ep)
+
+  list(APPEND ARROW_BUNDLED_STATIC_LIBS google-cloud-cpp::storage
+       google-cloud-cpp::common)
+endmacro()
+
+if(ARROW_WITH_GOOGLE_CLOUD_CPP)
+  resolve_dependency(google_cloud_cpp_storage)
+endif()
+
 #
 # HDFS thirdparty setup
 

diff --git a/cpp/thirdparty/versions.txt b/cpp/thirdparty/versions.txt
@@ -41,12 +41,16 @@ ARROW_BZIP2_BUILD_VERSION=1.0.8
 ARROW_BZIP2_BUILD_SHA256_CHECKSUM=ab5a03176ee106d3f0fa90e381da478ddae405918153cca248e682cd0c4a2269
 ARROW_CARES_BUILD_VERSION=1.17.1
 ARROW_CARES_BUILD_SHA256_CHECKSUM=d73dd0f6de824afd407ce10750ea081af47eba52b8a6cb307d220131ad93fc40
+ARROW_CRC32C_BUILD_VERSION=1.1.1
+ARROW_CRC32C_BUILD_SHA256_CHECKSUM=a6533f45b1670b5d59b38a514d82b09c6fb70cc1050467220216335e873074e8
 ARROW_GBENCHMARK_BUILD_VERSION=v1.5.2
 ARROW_GBENCHMARK_BUILD_SHA256_CHECKSUM=dccbdab796baa1043f04982147e67bb6e118fe610da2c65f88912d73987e700c
 ARROW_GFLAGS_BUILD_VERSION=v2.2.2
 ARROW_GFLAGS_BUILD_SHA256_CHECKSUM=34af2f15cf7367513b352bdcd2493ab14ce43692d2dcd9dfc499492966c64dcf
 ARROW_GLOG_BUILD_VERSION=v0.4.0
 ARROW_GLOG_BUILD_SHA256_CHECKSUM=f28359aeba12f30d73d9e4711ef356dc842886968112162bc73002645139c39c
+ARROW_GOOGLE_CLOUD_CPP_BUILD_VERSION=v1.31.1
+ARROW_GOOGLE_CLOUD_CPP_BUILD_SHA256_CHECKSUM=dc7cbf95b506a84b48cf71e0462985d262183edeaabdacaaee2109852394a609
 ARROW_GRPC_BUILD_VERSION=v1.35.0
 ARROW_GRPC_BUILD_SHA256_CHECKSUM=27dd2fc5c9809ddcde8eb6fa1fa278a3486566dfc28335fca13eb8df8bd3b958
 ARROW_GTEST_BUILD_VERSION=1.11.0
@@ -59,6 +63,8 @@ ARROW_LZ4_BUILD_SHA256_CHECKSUM=030644df4611007ff7dc962d981f390361e6c97a34e5cbc3
 # https://github.com/microsoft/mimalloc/issues/353
 ARROW_MIMALLOC_BUILD_VERSION=v1.7.2
 ARROW_MIMALLOC_BUILD_SHA256_CHECKSUM=b1912e354565a4b698410f7583c0f83934a6dbb3ade54ab7ddcb1569320936bd
+ARROW_NLOHMANN_JSON_BUILD_VERSION=v3.10.2
+ARROW_NLOHMANN_JSON_BUILD_SHA256_CHECKSUM=081ed0f9f89805c2d96335c3acfa993b39a0a5b4b4cef7edb68dd2210a13458c
 ARROW_ORC_BUILD_VERSION=1.7.0
 ARROW_ORC_BUILD_SHA256_CHECKSUM=45d6ba9149ffa2aaa168d61ab326f61181861c94529f26da3918a9aa2f801e39
 ARROW_PROTOBUF_BUILD_VERSION=v3.17.3
@@ -98,14 +104,17 @@ DEPENDENCIES=(
   "ARROW_BROTLI_URL brotli-${ARROW_BROTLI_BUILD_VERSION}.tar.gz https://github.com/google/brotli/archive/${ARROW_BROTLI_BUILD_VERSION}.tar.gz"
   "ARROW_BZIP2_URL bzip2-${ARROW_BZIP2_BUILD_VERSION}.tar.gz https://sourceware.org/pub/bzip2/bzip2-${ARROW_BZIP2_BUILD_VERSION}.tar.gz"
   "ARROW_CARES_URL cares-${ARROW_CARES_BUILD_VERSION}.tar.gz https://c-ares.haxx.se/download/c-ares-${ARROW_CARES_BUILD_VERSION}.tar.gz"
+  "ARROW_CRC32C_URL crc32c-${ARROW_CRC32C_BUILD_VERSION}.tar.gz https://github.com/google/crc32c/archive/refs/tags/${ARROW_CRC32C_BUILD_VERSION}.tar.gz
   "ARROW_GBENCHMARK_URL gbenchmark-${ARROW_GBENCHMARK_BUILD_VERSION}.tar.gz https://github.com/google/benchmark/archive/${ARROW_GBENCHMARK_BUILD_VERSION}.tar.gz"
   "ARROW_GFLAGS_URL gflags-${ARROW_GFLAGS_BUILD_VERSION}.tar.gz https://github.com/gflags/gflags/archive/${ARROW_GFLAGS_BUILD_VERSION}.tar.gz"
   "ARROW_GLOG_URL glog-${ARROW_GLOG_BUILD_VERSION}.tar.gz https://github.com/google/glog/archive/${ARROW_GLOG_BUILD_VERSION}.tar.gz"
+  "ARROW_GOOGLE_CLOUD_CPP_URL google-cloud-cpp-${ARROW_GOOGLE_CLOUD_CPP_BUILD_VERSION}.tar.gz https://github.com/googleapis/google-cloud-cpp/archive/refs/tags/${ARROW_GOOGLE_CLOUD_CPP_BUILD_VERSION}.tar.gz
   "ARROW_GRPC_URL grpc-${ARROW_GRPC_BUILD_VERSION}.tar.gz https://github.com/grpc/grpc/archive/${ARROW_GRPC_BUILD_VERSION}.tar.gz"
   "ARROW_GTEST_URL gtest-${ARROW_GTEST_BUILD_VERSION}.tar.gz https://github.com/google/googletest/archive/release-${ARROW_GTEST_BUILD_VERSION}.tar.gz"
   "ARROW_JEMALLOC_URL jemalloc-${ARROW_JEMALLOC_BUILD_VERSION}.tar.bz2 https://github.com/jemalloc/jemalloc/releases/download/${ARROW_JEMALLOC_BUILD_VERSION}/jemalloc-${ARROW_JEMALLOC_BUILD_VERSION}.tar.bz2"
   "ARROW_LZ4_URL lz4-${ARROW_LZ4_BUILD_VERSION}.tar.gz https://github.com/lz4/lz4/archive/${ARROW_LZ4_BUILD_VERSION}.tar.gz"
   "ARROW_MIMALLOC_URL mimalloc-${ARROW_MIMALLOC_BUILD_VERSION}.tar.gz https://github.com/microsoft/mimalloc/archive/${ARROW_MIMALLOC_BUILD_VERSION}.tar.gz"
+  "ARROW_NLOHMANN_JSON_URL nlohmann-json-${ARROW_NLOHMANN_JSON_BUILD_VERSION}.tar.gz https://github.com/nlohmann/json/archive/refs/tags/${ARROW_NLOHMANN_JSON_BUILD_VERSION}.tar.gz
   "ARROW_ORC_URL orc-${ARROW_ORC_BUILD_VERSION}.tar.gz https://github.com/apache/orc/archive/rel/release-${ARROW_ORC_BUILD_VERSION}.tar.gz"
   "ARROW_PROTOBUF_URL protobuf-${ARROW_PROTOBUF_BUILD_VERSION}.tar.gz https://github.com/google/protobuf/releases/download/${ARROW_PROTOBUF_BUILD_VERSION}/protobuf-all-${ARROW_PROTOBUF_BUILD_VERSION:1}.tar.gz"
   "ARROW_RAPIDJSON_URL rapidjson-${ARROW_RAPIDJSON_BUILD_VERSION}.tar.gz https://github.com/miloyip/rapidjson/archive/${ARROW_RAPIDJSON_BUILD_VERSION}.tar.gz"