pitrou · pitrou · Jan 13, 2025 · Jan 13, 2025 · Jan 13, 2025 · Jan 13, 2025
diff --git a/.asf.yaml b/.asf.yaml
@@ -23,7 +23,6 @@ github:
     - benibus
     - jbonofre
     - js8544
-    - laurentgo
     - vibhatha
     - ZhangHuiGui
 

diff --git a/.github/workflows/pr_review_trigger.yml b/.github/workflows/pr_review_trigger.yml
@@ -29,7 +29,7 @@ jobs:
     runs-on: ubuntu-latest
     steps:
       - name: "Upload PR review Payload"
-        uses: actions/upload-artifact@6f51ac03b9356f520e9adb1b1b7802705f340c2b # v4.5.0
+        uses: actions/upload-artifact@65c4c4a1ddee5b72f698fdd19549f0f0fb45cf08 # v4.6.0
         with:
           path: "${{ github.event_path }}"
           name: "pr_review_payload"
diff --git a/.github/workflows/r.yml b/.github/workflows/r.yml
@@ -177,7 +177,7 @@ jobs:
         if: always()
       - name: Save the test output
         if: always()
-        uses: actions/upload-artifact@6f51ac03b9356f520e9adb1b1b7802705f340c2b # v4.5.0
+        uses: actions/upload-artifact@65c4c4a1ddee5b72f698fdd19549f0f0fb45cf08 # v4.6.0
         with:
           name: test-output-${{ matrix.ubuntu }}-${{ matrix.r }}
           path: r/check/arrow.Rcheck/tests/testthat.Rout*
@@ -237,7 +237,7 @@ jobs:
         if: always()
       - name: Save the test output
         if: always()
-        uses: actions/upload-artifact@6f51ac03b9356f520e9adb1b1b7802705f340c2b # v4.5.0
+        uses: actions/upload-artifact@65c4c4a1ddee5b72f698fdd19549f0f0fb45cf08 # v4.6.0
         with:
           name: test-output-bundled
           path: r/check/arrow.Rcheck/tests/testthat.Rout*
@@ -299,7 +299,7 @@ jobs:
         # So that they're unique when multiple are downloaded in the next step
         shell: bash
         run: mv libarrow.zip libarrow-rtools${{ matrix.config.rtools }}-${{ matrix.config.arch }}.zip
-      - uses: actions/upload-artifact@6f51ac03b9356f520e9adb1b1b7802705f340c2b # v4.5.0
+      - uses: actions/upload-artifact@65c4c4a1ddee5b72f698fdd19549f0f0fb45cf08 # v4.6.0
         with:
           name: libarrow-rtools${{ matrix.config.rtools }}-${{ matrix.config.arch }}.zip
           path: libarrow-rtools${{ matrix.config.rtools }}-${{ matrix.config.arch }}.zip

diff --git a/c_glib/parquet-glib/arrow-file-writer.cpp b/c_glib/parquet-glib/arrow-file-writer.cpp
@@ -574,7 +574,6 @@ gparquet_arrow_file_writer_write_table(GParquetArrowFileWriter *writer,
 /**
  * gparquet_arrow_file_writer_new_row_group:
  * @writer: A #GParquetArrowFileWriter.
- * @chunk_size: The max number of rows in a row group.
  * @error: (nullable): Return location for a #GError or %NULL.
  *
  * Start a new row group.
@@ -584,13 +583,11 @@ gparquet_arrow_file_writer_write_table(GParquetArrowFileWriter *writer,
  * Since: 18.0.0
  */
 gboolean
-gparquet_arrow_file_writer_new_row_group(GParquetArrowFileWriter *writer,
-                                         gsize chunk_size,
-                                         GError **error)
+gparquet_arrow_file_writer_new_row_group(GParquetArrowFileWriter *writer, GError **error)
 {
   auto parquet_arrow_file_writer = gparquet_arrow_file_writer_get_raw(writer);
   return garrow::check(error,
-                       parquet_arrow_file_writer->NewRowGroup(chunk_size),
+                       parquet_arrow_file_writer->NewRowGroup(),
                        "[parquet][arrow][file-writer][new-row-group]");
 }
 

diff --git a/c_glib/parquet-glib/arrow-file-writer.h b/c_glib/parquet-glib/arrow-file-writer.h
@@ -135,9 +135,7 @@ gparquet_arrow_file_writer_write_table(GParquetArrowFileWriter *writer,
 
 GPARQUET_AVAILABLE_IN_18_0
 gboolean
-gparquet_arrow_file_writer_new_row_group(GParquetArrowFileWriter *writer,
-                                         gsize chunk_size,
-                                         GError **error);
+gparquet_arrow_file_writer_new_row_group(GParquetArrowFileWriter *writer, GError **error);
 
 GPARQUET_AVAILABLE_IN_18_0
 gboolean

diff --git a/c_glib/test/parquet/test-arrow-file-writer.rb b/c_glib/test/parquet/test-arrow-file-writer.rb
@@ -89,10 +89,10 @@ def test_write_table
   def test_write_chunked_array
     schema = build_schema("enabled" => :boolean)
     writer = Parquet::ArrowFileWriter.new(schema, @file.path)
-    writer.new_row_group(2)
+    writer.new_row_group
     chunked_array = Arrow::ChunkedArray.new([build_boolean_array([true, nil])])
     writer.write_chunked_array(chunked_array)
-    writer.new_row_group(1)
+    writer.new_row_group
     chunked_array = Arrow::ChunkedArray.new([build_boolean_array([false])])
     writer.write_chunked_array(chunked_array)
     writer.close

diff --git a/ci/conda_env_sphinx.txt b/ci/conda_env_sphinx.txt
@@ -30,9 +30,5 @@ sphinx-lint
 sphinxcontrib-jquery
 sphinxcontrib-mermaid
 sphinx==6.2
-# Requirement for doctest-cython
-# Needs upper pin of 0.3.0, see:
-# https://github.com/lgpage/pytest-cython/issues/67
-# With 0.3.* bug fix release, the pin can be removed
-pytest-cython==0.2.2
+pytest-cython
 pandas
diff --git a/ci/docker/debian-12-cpp.dockerfile b/ci/docker/debian-12-cpp.dockerfile
@@ -84,6 +84,7 @@ RUN apt-get update -y -q && \
         ninja-build \
         nlohmann-json3-dev \
         npm \
+        patch \
         pkg-config \
         protobuf-compiler-grpc \
         python3-dev \

diff --git a/ci/docker/ubuntu-20.04-cpp.dockerfile b/ci/docker/ubuntu-20.04-cpp.dockerfile
@@ -106,6 +106,7 @@ RUN apt-get update -y -q && \
         ninja-build \
         nlohmann-json3-dev \
         npm \
+        patch \
         pkg-config \
         protobuf-compiler \
         python3-dev \

diff --git a/ci/docker/ubuntu-22.04-cpp.dockerfile b/ci/docker/ubuntu-22.04-cpp.dockerfile
@@ -111,6 +111,7 @@ RUN apt-get update -y -q && \
         ninja-build \
         nlohmann-json3-dev \
         npm \
+        patch \
         pkg-config \
         protobuf-compiler \
         protobuf-compiler-grpc \

diff --git a/ci/docker/ubuntu-24.04-cpp.dockerfile b/ci/docker/ubuntu-24.04-cpp.dockerfile
@@ -111,6 +111,7 @@ RUN apt-get update -y -q && \
         ninja-build \
         nlohmann-json3-dev \
         npm \
+        patch \
         pkg-config \
         protobuf-compiler \
         protobuf-compiler-grpc \

diff --git a/cpp/cmake_modules/ThirdpartyToolchain.cmake b/cpp/cmake_modules/ThirdpartyToolchain.cmake
@@ -4573,11 +4573,16 @@ target_include_directories(arrow::hadoop INTERFACE "${HADOOP_HOME}/include")
 function(build_orc)
   message(STATUS "Building Apache ORC from source")
 
+  # Remove this and "patch" in "ci/docker/{debian,ubuntu}-*.dockerfile" once we have a patch for ORC 2.1.1
+  find_program(PATCH patch REQUIRED)
+  set(ORC_PATCH_COMMAND ${PATCH} -p1 -i ${CMAKE_CURRENT_LIST_DIR}/orc.diff)
+
   if(CMAKE_VERSION VERSION_GREATER_EQUAL 3.29)
     fetchcontent_declare(orc
                          ${FC_DECLARE_COMMON_OPTIONS}
                          URL ${ORC_SOURCE_URL}
-                         URL_HASH "SHA256=${ARROW_ORC_BUILD_SHA256_CHECKSUM}")
+                         URL_HASH "SHA256=${ARROW_ORC_BUILD_SHA256_CHECKSUM}"
+                         PATCH_COMMAND ${ORC_PATCH_COMMAND})
     prepare_fetchcontent()
 
     set(CMAKE_UNITY_BUILD FALSE)
@@ -4667,16 +4672,10 @@ function(build_orc)
         OFF
         CACHE BOOL "" FORCE)
 
-    # We can remove this with ORC 2.0.2 or later.
-    list(PREPEND CMAKE_MODULE_PATH
-         ${CMAKE_CURRENT_BINARY_DIR}/_deps/orc-src/cmake_modules)
-
     fetchcontent_makeavailable(orc)
 
     add_library(orc::orc INTERFACE IMPORTED)
     target_link_libraries(orc::orc INTERFACE orc)
-    target_include_directories(orc::orc INTERFACE "${orc_BINARY_DIR}/c++/include"
-                                                  "${orc_SOURCE_DIR}/c++/include")
 
     list(APPEND ARROW_BUNDLED_STATIC_LIBS orc)
   else()
@@ -4701,6 +4700,9 @@ function(build_orc)
     get_target_property(ORC_ZSTD_ROOT ${ARROW_ZSTD_LIBZSTD} INTERFACE_INCLUDE_DIRECTORIES)
     get_filename_component(ORC_ZSTD_ROOT "${ORC_ZSTD_ROOT}" DIRECTORY)
 
+    get_target_property(ORC_ZLIB_ROOT ZLIB::ZLIB INTERFACE_INCLUDE_DIRECTORIES)
+    get_filename_component(ORC_ZLIB_ROOT "${ORC_ZLIB_ROOT}" DIRECTORY)
+
     set(ORC_CMAKE_ARGS
         ${EP_COMMON_CMAKE_ARGS}
         "-DCMAKE_INSTALL_PREFIX=${ORC_PREFIX}"
@@ -4710,24 +4712,24 @@ function(build_orc)
         -DBUILD_TOOLS=OFF
         -DBUILD_CPP_TESTS=OFF
         -DINSTALL_VENDORED_LIBS=OFF
-        "-DLZ4_HOME=${ORC_LZ4_ROOT}"
         "-DPROTOBUF_EXECUTABLE=$<TARGET_FILE:${ARROW_PROTOBUF_PROTOC}>"
         "-DPROTOBUF_HOME=${ORC_PROTOBUF_ROOT}"
         "-DPROTOBUF_INCLUDE_DIR=$<TARGET_PROPERTY:${ARROW_PROTOBUF_LIBPROTOBUF},INTERFACE_INCLUDE_DIRECTORIES>"
         "-DPROTOBUF_LIBRARY=$<TARGET_FILE:${ARROW_PROTOBUF_LIBPROTOBUF}>"
         "-DPROTOC_LIBRARY=$<TARGET_FILE:${ARROW_PROTOBUF_LIBPROTOC}>"
         "-DSNAPPY_HOME=${ORC_SNAPPY_ROOT}"
         "-DSNAPPY_LIBRARY=$<TARGET_FILE:${Snappy_TARGET}>"
+        "-DLZ4_HOME=${ORC_LZ4_ROOT}"
         "-DLZ4_LIBRARY=$<TARGET_FILE:LZ4::lz4>"
         "-DLZ4_STATIC_LIB=$<TARGET_FILE:LZ4::lz4>"
         "-DLZ4_INCLUDE_DIR=${ORC_LZ4_ROOT}/include"
         "-DSNAPPY_INCLUDE_DIR=${ORC_SNAPPY_INCLUDE_DIR}"
         "-DZSTD_HOME=${ORC_ZSTD_ROOT}"
         "-DZSTD_INCLUDE_DIR=$<TARGET_PROPERTY:${ARROW_ZSTD_LIBZSTD},INTERFACE_INCLUDE_DIRECTORIES>"
-        "-DZSTD_LIBRARY=$<TARGET_FILE:${ARROW_ZSTD_LIBZSTD}>")
-    if(ZLIB_ROOT)
-      set(ORC_CMAKE_ARGS ${ORC_CMAKE_ARGS} "-DZLIB_HOME=${ZLIB_ROOT}")
-    endif()
+        "-DZSTD_LIBRARY=$<TARGET_FILE:${ARROW_ZSTD_LIBZSTD}>"
+        "-DZLIB_HOME=${ORC_ZLIB_ROOT}"
+        "-DZLIB_INCLUDE_DIR=$<TARGET_PROPERTY:ZLIB::ZLIB,INTERFACE_INCLUDE_DIRECTORIES>"
+        "-DZLIB_LIBRARY=$<TARGET_FILE:ZLIB::ZLIB>")
 
     # Work around CMake bug
     file(MAKE_DIRECTORY ${ORC_INCLUDE_DIR})
@@ -4743,7 +4745,8 @@ function(build_orc)
                                 ${ARROW_ZSTD_LIBZSTD}
                                 ${Snappy_TARGET}
                                 LZ4::lz4
-                                ZLIB::ZLIB)
+                                ZLIB::ZLIB
+                        PATCH_COMMAND ${ORC_PATCH_COMMAND})
     add_library(orc::orc STATIC IMPORTED)
     set_target_properties(orc::orc PROPERTIES IMPORTED_LOCATION "${ORC_STATIC_LIB}")
     target_include_directories(orc::orc BEFORE INTERFACE "${ORC_INCLUDE_DIR}")