GH-47329: [Packaging][Python][C++] Build and test PyArrow against Pyodide 314 (2026_0 ABI)

ci/docker/conda-python-emscripten.dockerfile

@@ -18,32 +18,39 @@
 ARG repo
 ARG arch
 ARG arch_short
-ARG python="3.12"
+ARG python="3.14"
 FROM --platform=linux/${arch} ${repo}:${arch_short}-conda-python-${python}
 
 ARG selenium_version="4.41.0"
-ARG pyodide_version="0.26.0"
+ARG pyodide_version="314.0.4"
+# pyodide-build is versioned independently of Pyodide
+ARG pyodide_build_version="0.39.0"
 ARG chrome_version="latest"
-ARG required_python_min="(3,12)"
-# fail if python version < 3.12
+ARG required_python_min="(3,14)"
+# fail if python version < 3.14
 RUN echo "check PYTHON>=${required_python_min}" && python -c "import sys;sys.exit(0 if sys.version_info>=${required_python_min} else 1)"
 
-# install selenium and recent pyodide-build and recent python
+RUN apt-get update -y -q && \
+ apt-get install -y -q --no-install-recommends \
+ libatomic1 && \
+ apt-get clean && \
+ rm -rf /var/lib/apt/lists/*
 
 # needs to be a login shell so ~/.profile is read
 SHELL ["/bin/bash", "--login", "-c", "-o", "pipefail"]
 
+# install selenium and recent pyodide-build and recent python
 RUN python -m pip install --no-cache-dir selenium==${selenium_version} && \
- python -m pip install --no-cache-dir --upgrade pyodide-build>=${pyodide_version}
+ python -m pip install --no-cache-dir --upgrade "pyodide-build>=${pyodide_build_version}"
 
 # install pyodide dist directory to /pyodide
 RUN pyodide_dist_url="https://github.com/pyodide/pyodide/releases/download/${pyodide_version}/pyodide-${pyodide_version}.tar.bz2" && \
  wget -q "${pyodide_dist_url}" -O- | tar -xj -C /
 
-# install node 20 (needed for async call support)
+# install node 24 (needed for async call support and JSPI)
 # and pthread-stubs for build, and unzip needed for chrome build to work
 # xz is needed by emsdk to extract node tarballs
-RUN conda install nodejs=20 unzip pthread-stubs make xz -c conda-forge
+RUN conda install nodejs=24 unzip pthread-stubs make xz -c conda-forge
 
 # install correct version of emscripten for this pyodide
 COPY ci/scripts/install_emscripten.sh /arrow/ci/scripts/

compose.yaml

@@ -981,10 +981,11 @@ services:
  arch_short: ${ARCH_SHORT}
  clang_tools: ${CLANG_TOOLS}
  llvm: ${LLVM}
- pyodide_version: "0.26.0"
+ pyodide_version: "314.0.4"
+ pyodide_build_version: "0.39.0"
  chrome_version: "latest"
  selenium_version: "4.41.0"
- required_python_min: "(3,12)"
+ required_python_min: "(3,14)"
  python: ${PYTHON}
  shm_size: *shm-size
  volumes: *ubuntu-volumes

cpp/cmake_modules/SetupCxxFlags.cmake

@@ -719,26 +719,35 @@ if(MSVC)
 endif()
 
 if(CMAKE_SYSTEM_NAME STREQUAL "Emscripten")
- # flags are:
+ # For the Pyodide 2026_0 ABI, the flags are:
  # 1) We force *everything* to build as position independent
- # 2) And with support for C++ exceptions
- set(CMAKE_C_FLAGS "${CMAKE_C_FLAGS} -fPIC -fexceptions")
+ # 2) We support C++ WASM exceptions, using the native WebAssembly
+ # exception handling proposal and WebAssembly-based setjmp/longjmp
+ set(CMAKE_C_FLAGS "${CMAKE_C_FLAGS} -fPIC -fwasm-exceptions -sSUPPORT_LONGJMP=wasm")
  # deprecated-literal-operator error is thrown in datetime (vendored lib in arrow)
  set(CMAKE_CXX_FLAGS
- "${CMAKE_CXX_FLAGS} -fPIC -fexceptions -Wno-error=deprecated-literal-operator")
+ "${CMAKE_CXX_FLAGS} -fPIC -fwasm-exceptions -sSUPPORT_LONGJMP=wasm -Wno-error=deprecated-literal-operator"
+ )
 
  # flags for creating shared libraries (only used in pyarrow, because
  # Emscripten builds libarrow as static)
  # flags are:
  # 1) Tell it to use JavaScript / WebAssembly 64 bit number support.
- # 2) Tell it to build with support for C++ exceptions
+ # 2) Tell it to build with support for C++ exceptions (see above)
  # 3) Skip linker flags error which happens with -soname parameter
- set(ARROW_EMSCRIPTEN_LINKER_FLAGS "-sWASM_BIGINT=1 -fexceptions -Wno-error=linkflags")
- set(CMAKE_SHARED_LIBRARY_CREATE_C_FLAGS
- "-sSIDE_MODULE=1 ${ARROW_EMSCRIPTEN_LINKER_FLAGS}")
- set(CMAKE_SHARED_LIBRARY_CREATE_CXX_FLAGS
- "-sSIDE_MODULE=1 ${ARROW_EMSCRIPTEN_LINKER_FLAGS}")
- set(CMAKE_SHARED_LINKER_FLAGS "-sSIDE_MODULE=1 ${ARROW_EMSCRIPTEN_LINKER_FLAGS}")
+ # 4) Record $ORIGIN as an RPATH, so that Pyodide can locate
+ # libarrow_python.so next to the extension modules needing it
+ # Note: the CMAKE_SHARED_LIBRARY_CREATE_<LANG>_FLAGS and
+ # CMAKE_SHARED_MODULE_CREATE_<LANG>_FLAGS variables are not used by the
+ # Ninja generator, so the flags must go through the linker-flags
+ # variables, which apply to SHARED (libarrow_python) and MODULE (Cython
+ # extension) targets respectively.
+ set(ARROW_EMSCRIPTEN_LINKER_FLAGS
+ "-sWASM_BIGINT=1 -fwasm-exceptions -sSUPPORT_LONGJMP=wasm -Wno-error=linkflags")
+ set(ARROW_EMSCRIPTEN_DYLIB_CREATE_FLAGS
+ "-sSIDE_MODULE=1 ${ARROW_EMSCRIPTEN_LINKER_FLAGS} '-Wl,-rpath,$ORIGIN'")
+ set(CMAKE_SHARED_LINKER_FLAGS "${ARROW_EMSCRIPTEN_DYLIB_CREATE_FLAGS}")
+ set(CMAKE_MODULE_LINKER_FLAGS "${ARROW_EMSCRIPTEN_DYLIB_CREATE_FLAGS}")
  if(ARROW_TESTING)
  # flags for building test executables for use in node
  if("${UPPERCASE_BUILD_TYPE}" STREQUAL "RELEASE")

cpp/cmake_modules/ThirdpartyToolchain.cmake

@@ -2183,6 +2183,10 @@ function(build_protobuf)
  set(PROTOBUF_HOST_CMAKE_ARGS
  "-DCMAKE_CXX_FLAGS="
  "-DCMAKE_C_FLAGS="
+ # Explicitly set the C++ standard for the host build. Without this,
+ # Abseil may be compiled with the host compiler's default standard,
+ # which can be older than C++17.
+ "-DCMAKE_CXX_STANDARD=${CMAKE_CXX_STANDARD}"
  "-DCMAKE_INSTALL_PREFIX=${PROTOBUF_HOST_PREFIX}"
  -Dprotobuf_BUILD_TESTS=OFF
  -Dprotobuf_DEBUG_POSTFIX=

dev/tasks/tasks.yml

@@ -521,7 +521,7 @@ tasks:
  params:
  env:
  UBUNTU: 22.04
- PYTHON: 3.12
+ PYTHON: 3.14
  image: conda-python-emscripten
 
  test-conda-python-3.11-hypothesis:

docs/source/developers/cpp/emscripten.rst

@@ -34,9 +34,9 @@ activate it using the commands below (see https://emscripten.org/docs/getting_st
  git clone https://github.com/emscripten-core/emsdk.git
  cd emsdk
  # replace <version> with the desired EMSDK version.
- # e.g. for Pyodide 0.26, you need EMSDK version 3.1.58
+ # e.g. for Pyodide 314.0.x, you need EMSDK version 5.0.3
  # the versions can be found in the Makefile.envs file in the Pyodide repo:
- # https://github.com/pyodide/pyodide/blob/10b484cfe427e076c929a55dc35cfff01ea8d3bc/Makefile.envs
+ # https://github.com/pyodide/pyodide/blob/314.0.4/Makefile.envs
  ./emsdk install <version>
  ./emsdk activate <version>
  source ./emsdk_env.sh

python/CMakeLists.txt

@@ -111,7 +111,9 @@ if("$ENV{CMAKE_EXPORT_COMPILE_COMMANDS}" STREQUAL "1")
  set(CMAKE_EXPORT_COMPILE_COMMANDS 1)
 endif()
 
-if(UNIX)
+# On Emscripten, the equivalent $ORIGIN runtime path is added
+# through the linker flags in SetupCxxFlags.cmake instead
+if(UNIX AND NOT CMAKE_SYSTEM_NAME STREQUAL "Emscripten")
  set(CMAKE_BUILD_WITH_INSTALL_RPATH TRUE)
  # In the event that we are bundling the shared libraries (e.g. in a
  # manylinux1 wheel), we need to set the RPATH of the extensions to the

python/scripts/run_emscripten_tests.py

@@ -148,8 +148,14 @@ class NodeDriver:
  import subprocess
 
  def __init__(self, hostname, port):
+ # handle different syntax for util-linux script and macOS BSD-style script
+ if sys.platform == "darwin":
+ script_command = [shutil.which("script"), "-q", "/dev/null",
+ shutil.which("node")]
+ else:
+ script_command = [shutil.which("script"), "-c", shutil.which("node")]
  self.process = subprocess.Popen(
-[shutil.which("script"), "-c", shutil.which("node")],
+script_command,
  stdin=subprocess.PIPE,
  shell=False,
  bufsize=0,
@@ -320,6 +326,7 @@ def _load_pyarrow_in_runner(driver, wheel_name):
 if "pyarrow" not in sys.modules:
  await micropip.install("hypothesis")
  import pyodide_js as pjs
+ await pjs.loadPackage("packaging")
  await pjs.loadPackage("tzdata")
  await pjs.loadPackage("numpy")
  await pjs.loadPackage("pandas")

ci/docker/conda-python-emscripten.dockerfile

@@ -18,32 +18,39 @@
 ARG repo
 ARG arch
 ARG arch_short
-ARG python="3.12"
+ARG python="3.14"
 FROM --platform=linux/${arch} ${repo}:${arch_short}-conda-python-${python}
 
 ARG selenium_version="4.41.0"
-ARG pyodide_version="0.26.0"
+ARG pyodide_version="314.0.4"
+# pyodide-build is versioned independently of Pyodide
+ARG pyodide_build_version="0.39.0"
 ARG chrome_version="latest"
-ARG required_python_min="(3,12)"
-# fail if python version < 3.12
+ARG required_python_min="(3,14)"
+# fail if python version < 3.14
 RUN echo "check PYTHON>=${required_python_min}" && python -c "import sys;sys.exit(0 if sys.version_info>=${required_python_min} else 1)"
 
-# install selenium and recent pyodide-build and recent python
+RUN apt-get update -y -q && \
+ apt-get install -y -q --no-install-recommends \
+ libatomic1 && \
+ apt-get clean && \
+ rm -rf /var/lib/apt/lists/*
 
 # needs to be a login shell so ~/.profile is read
 SHELL ["/bin/bash", "--login", "-c", "-o", "pipefail"]
 
+# install selenium and recent pyodide-build and recent python
 RUN python -m pip install --no-cache-dir selenium==${selenium_version} && \
- python -m pip install --no-cache-dir --upgrade pyodide-build>=${pyodide_version}
+ python -m pip install --no-cache-dir --upgrade "pyodide-build>=${pyodide_build_version}"
 
 # install pyodide dist directory to /pyodide
 RUN pyodide_dist_url="https://github.com/pyodide/pyodide/releases/download/${pyodide_version}/pyodide-${pyodide_version}.tar.bz2" && \
  wget -q "${pyodide_dist_url}" -O- | tar -xj -C /
 
-# install node 20 (needed for async call support)
+# install node 24 (needed for async call support and JSPI)
 # and pthread-stubs for build, and unzip needed for chrome build to work
 # xz is needed by emsdk to extract node tarballs
-RUN conda install nodejs=20 unzip pthread-stubs make xz -c conda-forge
+RUN conda install nodejs=24 unzip pthread-stubs make xz -c conda-forge
 
 # install correct version of emscripten for this pyodide
 COPY ci/scripts/install_emscripten.sh /arrow/ci/scripts/

compose.yaml

@@ -981,10 +981,11 @@ services:
  arch_short: ${ARCH_SHORT}
  clang_tools: ${CLANG_TOOLS}
  llvm: ${LLVM}
- pyodide_version: "0.26.0"
+ pyodide_version: "314.0.4"
+ pyodide_build_version: "0.39.0"
  chrome_version: "latest"
  selenium_version: "4.41.0"
- required_python_min: "(3,12)"
+ required_python_min: "(3,14)"
  python: ${PYTHON}
  shm_size: *shm-size
  volumes: *ubuntu-volumes

cpp/cmake_modules/SetupCxxFlags.cmake

@@ -719,26 +719,35 @@ if(MSVC)
 endif()
 
 if(CMAKE_SYSTEM_NAME STREQUAL "Emscripten")
- # flags are:
+ # For the Pyodide 2026_0 ABI, the flags are:
  # 1) We force *everything* to build as position independent
- # 2) And with support for C++ exceptions
- set(CMAKE_C_FLAGS "${CMAKE_C_FLAGS} -fPIC -fexceptions")
+ # 2) We support C++ WASM exceptions, using the native WebAssembly
+ # exception handling proposal and WebAssembly-based setjmp/longjmp
+ set(CMAKE_C_FLAGS "${CMAKE_C_FLAGS} -fPIC -fwasm-exceptions -sSUPPORT_LONGJMP=wasm")
  # deprecated-literal-operator error is thrown in datetime (vendored lib in arrow)
  set(CMAKE_CXX_FLAGS
- "${CMAKE_CXX_FLAGS} -fPIC -fexceptions -Wno-error=deprecated-literal-operator")
+ "${CMAKE_CXX_FLAGS} -fPIC -fwasm-exceptions -sSUPPORT_LONGJMP=wasm -Wno-error=deprecated-literal-operator"
+ )
 
  # flags for creating shared libraries (only used in pyarrow, because
  # Emscripten builds libarrow as static)
  # flags are:
  # 1) Tell it to use JavaScript / WebAssembly 64 bit number support.
- # 2) Tell it to build with support for C++ exceptions
+ # 2) Tell it to build with support for C++ exceptions (see above)
  # 3) Skip linker flags error which happens with -soname parameter
- set(ARROW_EMSCRIPTEN_LINKER_FLAGS "-sWASM_BIGINT=1 -fexceptions -Wno-error=linkflags")
- set(CMAKE_SHARED_LIBRARY_CREATE_C_FLAGS
- "-sSIDE_MODULE=1 ${ARROW_EMSCRIPTEN_LINKER_FLAGS}")
- set(CMAKE_SHARED_LIBRARY_CREATE_CXX_FLAGS
- "-sSIDE_MODULE=1 ${ARROW_EMSCRIPTEN_LINKER_FLAGS}")
- set(CMAKE_SHARED_LINKER_FLAGS "-sSIDE_MODULE=1 ${ARROW_EMSCRIPTEN_LINKER_FLAGS}")
+ # 4) Record $ORIGIN as an RPATH, so that Pyodide can locate
+ # libarrow_python.so next to the extension modules needing it
+ # Note: the CMAKE_SHARED_LIBRARY_CREATE_<LANG>_FLAGS and
+ # CMAKE_SHARED_MODULE_CREATE_<LANG>_FLAGS variables are not used by the
+ # Ninja generator, so the flags must go through the linker-flags
+ # variables, which apply to SHARED (libarrow_python) and MODULE (Cython
+ # extension) targets respectively.
+ set(ARROW_EMSCRIPTEN_LINKER_FLAGS
+ "-sWASM_BIGINT=1 -fwasm-exceptions -sSUPPORT_LONGJMP=wasm -Wno-error=linkflags")
+ set(ARROW_EMSCRIPTEN_DYLIB_CREATE_FLAGS
+ "-sSIDE_MODULE=1 ${ARROW_EMSCRIPTEN_LINKER_FLAGS} '-Wl,-rpath,$ORIGIN'")
+ set(CMAKE_SHARED_LINKER_FLAGS "${ARROW_EMSCRIPTEN_DYLIB_CREATE_FLAGS}")
+ set(CMAKE_MODULE_LINKER_FLAGS "${ARROW_EMSCRIPTEN_DYLIB_CREATE_FLAGS}")
  if(ARROW_TESTING)
  # flags for building test executables for use in node
  if("${UPPERCASE_BUILD_TYPE}" STREQUAL "RELEASE")

cpp/cmake_modules/ThirdpartyToolchain.cmake

@@ -2183,6 +2183,10 @@ function(build_protobuf)
  set(PROTOBUF_HOST_CMAKE_ARGS
  "-DCMAKE_CXX_FLAGS="
  "-DCMAKE_C_FLAGS="
+ # Explicitly set the C++ standard for the host build. Without this,
+ # Abseil may be compiled with the host compiler's default standard,
+ # which can be older than C++17.
+ "-DCMAKE_CXX_STANDARD=${CMAKE_CXX_STANDARD}"
  "-DCMAKE_INSTALL_PREFIX=${PROTOBUF_HOST_PREFIX}"
  -Dprotobuf_BUILD_TESTS=OFF
  -Dprotobuf_DEBUG_POSTFIX=

dev/tasks/tasks.yml

@@ -521,7 +521,7 @@ tasks:
  params:
  env:
  UBUNTU: 22.04
- PYTHON: 3.12
+ PYTHON: 3.14
  image: conda-python-emscripten
 
  test-conda-python-3.11-hypothesis:

docs/source/developers/cpp/emscripten.rst

@@ -34,9 +34,9 @@ activate it using the commands below (see https://emscripten.org/docs/getting_st
  git clone https://github.com/emscripten-core/emsdk.git
  cd emsdk
  # replace <version> with the desired EMSDK version.
- # e.g. for Pyodide 0.26, you need EMSDK version 3.1.58
+ # e.g. for Pyodide 314.0.x, you need EMSDK version 5.0.3
  # the versions can be found in the Makefile.envs file in the Pyodide repo:
- # https://github.com/pyodide/pyodide/blob/10b484cfe427e076c929a55dc35cfff01ea8d3bc/Makefile.envs
+ # https://github.com/pyodide/pyodide/blob/314.0.4/Makefile.envs
  ./emsdk install <version>
  ./emsdk activate <version>
  source ./emsdk_env.sh

python/CMakeLists.txt

@@ -111,7 +111,9 @@ if("$ENV{CMAKE_EXPORT_COMPILE_COMMANDS}" STREQUAL "1")
  set(CMAKE_EXPORT_COMPILE_COMMANDS 1)
 endif()
 
-if(UNIX)
+# On Emscripten, the equivalent $ORIGIN runtime path is added
+# through the linker flags in SetupCxxFlags.cmake instead
+if(UNIX AND NOT CMAKE_SYSTEM_NAME STREQUAL "Emscripten")
  set(CMAKE_BUILD_WITH_INSTALL_RPATH TRUE)
  # In the event that we are bundling the shared libraries (e.g. in a
  # manylinux1 wheel), we need to set the RPATH of the extensions to the

python/scripts/run_emscripten_tests.py

@@ -148,8 +148,14 @@ class NodeDriver:
  import subprocess
 
  def __init__(self, hostname, port):
+ # handle different syntax for util-linux script and macOS BSD-style script
+ if sys.platform == "darwin":
+ script_command = [shutil.which("script"), "-q", "/dev/null",
+ shutil.which("node")]
+ else:
+ script_command = [shutil.which("script"), "-c", shutil.which("node")]
  self.process = subprocess.Popen(
-[shutil.which("script"), "-c", shutil.which("node")],
+script_command,
  stdin=subprocess.PIPE,
  shell=False,
  bufsize=0,
@@ -320,6 +326,7 @@ def _load_pyarrow_in_runner(driver, wheel_name):
 if "pyarrow" not in sys.modules:
  await micropip.install("hypothesis")
  import pyodide_js as pjs
+ await pjs.loadPackage("packaging")
  await pjs.loadPackage("tzdata")
  await pjs.loadPackage("numpy")
  await pjs.loadPackage("pandas")

ci/docker/conda-python-emscripten.dockerfile

@@ -18,32 +18,39 @@
 ARG repo
 ARG arch
 ARG arch_short
-ARG python="3.12"
+ARG python="3.14"
 FROM --platform=linux/${arch} ${repo}:${arch_short}-conda-python-${python}
 
 ARG selenium_version="4.41.0"
-ARG pyodide_version="0.26.0"
+ARG pyodide_version="314.0.4"
+# pyodide-build is versioned independently of Pyodide
+ARG pyodide_build_version="0.39.0"
 ARG chrome_version="latest"
-ARG required_python_min="(3,12)"
-# fail if python version < 3.12
+ARG required_python_min="(3,14)"
+# fail if python version < 3.14
 RUN echo "check PYTHON>=${required_python_min}" && python -c "import sys;sys.exit(0 if sys.version_info>=${required_python_min} else 1)"
 
-# install selenium and recent pyodide-build and recent python
+RUN apt-get update -y -q && \
+ apt-get install -y -q --no-install-recommends \
+ libatomic1 && \
+ apt-get clean && \
+ rm -rf /var/lib/apt/lists/*
 
 # needs to be a login shell so ~/.profile is read
 SHELL ["/bin/bash", "--login", "-c", "-o", "pipefail"]
 
+# install selenium and recent pyodide-build and recent python
 RUN python -m pip install --no-cache-dir selenium==${selenium_version} && \
- python -m pip install --no-cache-dir --upgrade pyodide-build>=${pyodide_version}
+ python -m pip install --no-cache-dir --upgrade "pyodide-build>=${pyodide_build_version}"
 
 # install pyodide dist directory to /pyodide
 RUN pyodide_dist_url="https://github.com/pyodide/pyodide/releases/download/${pyodide_version}/pyodide-${pyodide_version}.tar.bz2" && \
  wget -q "${pyodide_dist_url}" -O- | tar -xj -C /
 
-# install node 20 (needed for async call support)
+# install node 24 (needed for async call support and JSPI)
 # and pthread-stubs for build, and unzip needed for chrome build to work
 # xz is needed by emsdk to extract node tarballs
-RUN conda install nodejs=20 unzip pthread-stubs make xz -c conda-forge
+RUN conda install nodejs=24 unzip pthread-stubs make xz -c conda-forge
 
 # install correct version of emscripten for this pyodide
 COPY ci/scripts/install_emscripten.sh /arrow/ci/scripts/

compose.yaml

@@ -981,10 +981,11 @@ services:
  arch_short: ${ARCH_SHORT}
  clang_tools: ${CLANG_TOOLS}
  llvm: ${LLVM}
- pyodide_version: "0.26.0"
+ pyodide_version: "314.0.4"
+ pyodide_build_version: "0.39.0"
  chrome_version: "latest"
  selenium_version: "4.41.0"
- required_python_min: "(3,12)"
+ required_python_min: "(3,14)"
  python: ${PYTHON}
  shm_size: *shm-size
  volumes: *ubuntu-volumes

cpp/cmake_modules/SetupCxxFlags.cmake

@@ -719,26 +719,35 @@ if(MSVC)
 endif()
 
 if(CMAKE_SYSTEM_NAME STREQUAL "Emscripten")
- # flags are:
+ # For the Pyodide 2026_0 ABI, the flags are:
  # 1) We force *everything* to build as position independent
- # 2) And with support for C++ exceptions
- set(CMAKE_C_FLAGS "${CMAKE_C_FLAGS} -fPIC -fexceptions")
+ # 2) We support C++ WASM exceptions, using the native WebAssembly
+ # exception handling proposal and WebAssembly-based setjmp/longjmp
+ set(CMAKE_C_FLAGS "${CMAKE_C_FLAGS} -fPIC -fwasm-exceptions -sSUPPORT_LONGJMP=wasm")
  # deprecated-literal-operator error is thrown in datetime (vendored lib in arrow)
  set(CMAKE_CXX_FLAGS
- "${CMAKE_CXX_FLAGS} -fPIC -fexceptions -Wno-error=deprecated-literal-operator")
+ "${CMAKE_CXX_FLAGS} -fPIC -fwasm-exceptions -sSUPPORT_LONGJMP=wasm -Wno-error=deprecated-literal-operator"
+ )
 
  # flags for creating shared libraries (only used in pyarrow, because
  # Emscripten builds libarrow as static)
  # flags are:
  # 1) Tell it to use JavaScript / WebAssembly 64 bit number support.
- # 2) Tell it to build with support for C++ exceptions
+ # 2) Tell it to build with support for C++ exceptions (see above)
  # 3) Skip linker flags error which happens with -soname parameter
- set(ARROW_EMSCRIPTEN_LINKER_FLAGS "-sWASM_BIGINT=1 -fexceptions -Wno-error=linkflags")
- set(CMAKE_SHARED_LIBRARY_CREATE_C_FLAGS
- "-sSIDE_MODULE=1 ${ARROW_EMSCRIPTEN_LINKER_FLAGS}")
- set(CMAKE_SHARED_LIBRARY_CREATE_CXX_FLAGS
- "-sSIDE_MODULE=1 ${ARROW_EMSCRIPTEN_LINKER_FLAGS}")
- set(CMAKE_SHARED_LINKER_FLAGS "-sSIDE_MODULE=1 ${ARROW_EMSCRIPTEN_LINKER_FLAGS}")
+ # 4) Record $ORIGIN as an RPATH, so that Pyodide can locate
+ # libarrow_python.so next to the extension modules needing it
+ # Note: the CMAKE_SHARED_LIBRARY_CREATE_<LANG>_FLAGS and
+ # CMAKE_SHARED_MODULE_CREATE_<LANG>_FLAGS variables are not used by the
+ # Ninja generator, so the flags must go through the linker-flags
+ # variables, which apply to SHARED (libarrow_python) and MODULE (Cython
+ # extension) targets respectively.
+ set(ARROW_EMSCRIPTEN_LINKER_FLAGS
+ "-sWASM_BIGINT=1 -fwasm-exceptions -sSUPPORT_LONGJMP=wasm -Wno-error=linkflags")
+ set(ARROW_EMSCRIPTEN_DYLIB_CREATE_FLAGS
+ "-sSIDE_MODULE=1 ${ARROW_EMSCRIPTEN_LINKER_FLAGS} '-Wl,-rpath,$ORIGIN'")
+ set(CMAKE_SHARED_LINKER_FLAGS "${ARROW_EMSCRIPTEN_DYLIB_CREATE_FLAGS}")
+ set(CMAKE_MODULE_LINKER_FLAGS "${ARROW_EMSCRIPTEN_DYLIB_CREATE_FLAGS}")
  if(ARROW_TESTING)
  # flags for building test executables for use in node
  if("${UPPERCASE_BUILD_TYPE}" STREQUAL "RELEASE")

cpp/cmake_modules/ThirdpartyToolchain.cmake

@@ -2183,6 +2183,10 @@ function(build_protobuf)
  set(PROTOBUF_HOST_CMAKE_ARGS
  "-DCMAKE_CXX_FLAGS="
  "-DCMAKE_C_FLAGS="
+ # Explicitly set the C++ standard for the host build. Without this,
+ # Abseil may be compiled with the host compiler's default standard,
+ # which can be older than C++17.
+ "-DCMAKE_CXX_STANDARD=${CMAKE_CXX_STANDARD}"
  "-DCMAKE_INSTALL_PREFIX=${PROTOBUF_HOST_PREFIX}"
  -Dprotobuf_BUILD_TESTS=OFF
  -Dprotobuf_DEBUG_POSTFIX=

dev/tasks/tasks.yml

@@ -521,7 +521,7 @@ tasks:
  params:
  env:
  UBUNTU: 22.04
- PYTHON: 3.12
+ PYTHON: 3.14
  image: conda-python-emscripten
 
  test-conda-python-3.11-hypothesis:

docs/source/developers/cpp/emscripten.rst

@@ -34,9 +34,9 @@ activate it using the commands below (see https://emscripten.org/docs/getting_st
  git clone https://github.com/emscripten-core/emsdk.git
  cd emsdk
  # replace <version> with the desired EMSDK version.
- # e.g. for Pyodide 0.26, you need EMSDK version 3.1.58
+ # e.g. for Pyodide 314.0.x, you need EMSDK version 5.0.3
  # the versions can be found in the Makefile.envs file in the Pyodide repo:
- # https://github.com/pyodide/pyodide/blob/10b484cfe427e076c929a55dc35cfff01ea8d3bc/Makefile.envs
+ # https://github.com/pyodide/pyodide/blob/314.0.4/Makefile.envs
  ./emsdk install <version>
  ./emsdk activate <version>
  source ./emsdk_env.sh

python/CMakeLists.txt

@@ -111,7 +111,9 @@ if("$ENV{CMAKE_EXPORT_COMPILE_COMMANDS}" STREQUAL "1")
  set(CMAKE_EXPORT_COMPILE_COMMANDS 1)
 endif()
 
-if(UNIX)
+# On Emscripten, the equivalent $ORIGIN runtime path is added
+# through the linker flags in SetupCxxFlags.cmake instead
+if(UNIX AND NOT CMAKE_SYSTEM_NAME STREQUAL "Emscripten")
  set(CMAKE_BUILD_WITH_INSTALL_RPATH TRUE)
  # In the event that we are bundling the shared libraries (e.g. in a
  # manylinux1 wheel), we need to set the RPATH of the extensions to the

python/scripts/run_emscripten_tests.py

@@ -148,8 +148,14 @@ class NodeDriver:
  import subprocess
 
  def __init__(self, hostname, port):
+ # handle different syntax for util-linux script and macOS BSD-style script
+ if sys.platform == "darwin":
+ script_command = [shutil.which("script"), "-q", "/dev/null",
+ shutil.which("node")]
+ else:
+ script_command = [shutil.which("script"), "-c", shutil.which("node")]
  self.process = subprocess.Popen(
-[shutil.which("script"), "-c", shutil.which("node")],
+script_command,
  stdin=subprocess.PIPE,
  shell=False,
  bufsize=0,
@@ -320,6 +326,7 @@ def _load_pyarrow_in_runner(driver, wheel_name):
 if "pyarrow" not in sys.modules:
  await micropip.install("hypothesis")
  import pyodide_js as pjs
+ await pjs.loadPackage("packaging")
  await pjs.loadPackage("tzdata")
  await pjs.loadPackage("numpy")
  await pjs.loadPackage("pandas")

ci/docker/conda-python-emscripten.dockerfile

@@ -18,32 +18,39 @@
 ARG repo
 ARG arch
 ARG arch_short
-ARG python="3.12"
+ARG python="3.14"
 FROM --platform=linux/${arch} ${repo}:${arch_short}-conda-python-${python}
 
 ARG selenium_version="4.41.0"
-ARG pyodide_version="0.26.0"
+ARG pyodide_version="314.0.4"
+# pyodide-build is versioned independently of Pyodide
+ARG pyodide_build_version="0.39.0"
 ARG chrome_version="latest"
-ARG required_python_min="(3,12)"
-# fail if python version < 3.12
+ARG required_python_min="(3,14)"
+# fail if python version < 3.14
 RUN echo "check PYTHON>=${required_python_min}" && python -c "import sys;sys.exit(0 if sys.version_info>=${required_python_min} else 1)"
 
-# install selenium and recent pyodide-build and recent python
+RUN apt-get update -y -q && \
+ apt-get install -y -q --no-install-recommends \
+ libatomic1 && \
+ apt-get clean && \
+ rm -rf /var/lib/apt/lists/*
 
 # needs to be a login shell so ~/.profile is read
 SHELL ["/bin/bash", "--login", "-c", "-o", "pipefail"]
 
+# install selenium and recent pyodide-build and recent python
 RUN python -m pip install --no-cache-dir selenium==${selenium_version} && \
- python -m pip install --no-cache-dir --upgrade pyodide-build>=${pyodide_version}
+ python -m pip install --no-cache-dir --upgrade "pyodide-build>=${pyodide_build_version}"
 
 # install pyodide dist directory to /pyodide
 RUN pyodide_dist_url="https://github.com/pyodide/pyodide/releases/download/${pyodide_version}/pyodide-${pyodide_version}.tar.bz2" && \
  wget -q "${pyodide_dist_url}" -O- | tar -xj -C /
 
-# install node 20 (needed for async call support)
+# install node 24 (needed for async call support and JSPI)
 # and pthread-stubs for build, and unzip needed for chrome build to work
 # xz is needed by emsdk to extract node tarballs
-RUN conda install nodejs=20 unzip pthread-stubs make xz -c conda-forge
+RUN conda install nodejs=24 unzip pthread-stubs make xz -c conda-forge
 
 # install correct version of emscripten for this pyodide
 COPY ci/scripts/install_emscripten.sh /arrow/ci/scripts/

compose.yaml

@@ -981,10 +981,11 @@ services:
  arch_short: ${ARCH_SHORT}
  clang_tools: ${CLANG_TOOLS}
  llvm: ${LLVM}
- pyodide_version: "0.26.0"
+ pyodide_version: "314.0.4"
+ pyodide_build_version: "0.39.0"
  chrome_version: "latest"
  selenium_version: "4.41.0"
- required_python_min: "(3,12)"
+ required_python_min: "(3,14)"
  python: ${PYTHON}
  shm_size: *shm-size
  volumes: *ubuntu-volumes

cpp/cmake_modules/SetupCxxFlags.cmake

@@ -719,26 +719,35 @@ if(MSVC)
 endif()
 
 if(CMAKE_SYSTEM_NAME STREQUAL "Emscripten")
- # flags are:
+ # For the Pyodide 2026_0 ABI, the flags are:
  # 1) We force *everything* to build as position independent
- # 2) And with support for C++ exceptions
- set(CMAKE_C_FLAGS "${CMAKE_C_FLAGS} -fPIC -fexceptions")
+ # 2) We support C++ WASM exceptions, using the native WebAssembly
+ # exception handling proposal and WebAssembly-based setjmp/longjmp
+ set(CMAKE_C_FLAGS "${CMAKE_C_FLAGS} -fPIC -fwasm-exceptions -sSUPPORT_LONGJMP=wasm")
  # deprecated-literal-operator error is thrown in datetime (vendored lib in arrow)
  set(CMAKE_CXX_FLAGS
- "${CMAKE_CXX_FLAGS} -fPIC -fexceptions -Wno-error=deprecated-literal-operator")
+ "${CMAKE_CXX_FLAGS} -fPIC -fwasm-exceptions -sSUPPORT_LONGJMP=wasm -Wno-error=deprecated-literal-operator"
+ )
 
  # flags for creating shared libraries (only used in pyarrow, because
  # Emscripten builds libarrow as static)
  # flags are:
  # 1) Tell it to use JavaScript / WebAssembly 64 bit number support.
- # 2) Tell it to build with support for C++ exceptions
+ # 2) Tell it to build with support for C++ exceptions (see above)
  # 3) Skip linker flags error which happens with -soname parameter
- set(ARROW_EMSCRIPTEN_LINKER_FLAGS "-sWASM_BIGINT=1 -fexceptions -Wno-error=linkflags")
- set(CMAKE_SHARED_LIBRARY_CREATE_C_FLAGS
- "-sSIDE_MODULE=1 ${ARROW_EMSCRIPTEN_LINKER_FLAGS}")
- set(CMAKE_SHARED_LIBRARY_CREATE_CXX_FLAGS
- "-sSIDE_MODULE=1 ${ARROW_EMSCRIPTEN_LINKER_FLAGS}")
- set(CMAKE_SHARED_LINKER_FLAGS "-sSIDE_MODULE=1 ${ARROW_EMSCRIPTEN_LINKER_FLAGS}")
+ # 4) Record $ORIGIN as an RPATH, so that Pyodide can locate
+ # libarrow_python.so next to the extension modules needing it
+ # Note: the CMAKE_SHARED_LIBRARY_CREATE_<LANG>_FLAGS and
+ # CMAKE_SHARED_MODULE_CREATE_<LANG>_FLAGS variables are not used by the
+ # Ninja generator, so the flags must go through the linker-flags
+ # variables, which apply to SHARED (libarrow_python) and MODULE (Cython
+ # extension) targets respectively.
+ set(ARROW_EMSCRIPTEN_LINKER_FLAGS
+ "-sWASM_BIGINT=1 -fwasm-exceptions -sSUPPORT_LONGJMP=wasm -Wno-error=linkflags")
+ set(ARROW_EMSCRIPTEN_DYLIB_CREATE_FLAGS
+ "-sSIDE_MODULE=1 ${ARROW_EMSCRIPTEN_LINKER_FLAGS} '-Wl,-rpath,$ORIGIN'")
+ set(CMAKE_SHARED_LINKER_FLAGS "${ARROW_EMSCRIPTEN_DYLIB_CREATE_FLAGS}")
+ set(CMAKE_MODULE_LINKER_FLAGS "${ARROW_EMSCRIPTEN_DYLIB_CREATE_FLAGS}")
  if(ARROW_TESTING)
  # flags for building test executables for use in node
  if("${UPPERCASE_BUILD_TYPE}" STREQUAL "RELEASE")

cpp/cmake_modules/ThirdpartyToolchain.cmake

@@ -2183,6 +2183,10 @@ function(build_protobuf)
  set(PROTOBUF_HOST_CMAKE_ARGS
  "-DCMAKE_CXX_FLAGS="
  "-DCMAKE_C_FLAGS="
+ # Explicitly set the C++ standard for the host build. Without this,
+ # Abseil may be compiled with the host compiler's default standard,
+ # which can be older than C++17.
+ "-DCMAKE_CXX_STANDARD=${CMAKE_CXX_STANDARD}"
  "-DCMAKE_INSTALL_PREFIX=${PROTOBUF_HOST_PREFIX}"
  -Dprotobuf_BUILD_TESTS=OFF
  -Dprotobuf_DEBUG_POSTFIX=

dev/tasks/tasks.yml

@@ -521,7 +521,7 @@ tasks:
  params:
  env:
  UBUNTU: 22.04
- PYTHON: 3.12
+ PYTHON: 3.14
  image: conda-python-emscripten
 
  test-conda-python-3.11-hypothesis:

docs/source/developers/cpp/emscripten.rst

@@ -34,9 +34,9 @@ activate it using the commands below (see https://emscripten.org/docs/getting_st
  git clone https://github.com/emscripten-core/emsdk.git
  cd emsdk
  # replace <version> with the desired EMSDK version.
- # e.g. for Pyodide 0.26, you need EMSDK version 3.1.58
+ # e.g. for Pyodide 314.0.x, you need EMSDK version 5.0.3
  # the versions can be found in the Makefile.envs file in the Pyodide repo:
- # https://github.com/pyodide/pyodide/blob/10b484cfe427e076c929a55dc35cfff01ea8d3bc/Makefile.envs
+ # https://github.com/pyodide/pyodide/blob/314.0.4/Makefile.envs
  ./emsdk install <version>
  ./emsdk activate <version>
  source ./emsdk_env.sh

python/CMakeLists.txt

@@ -111,7 +111,9 @@ if("$ENV{CMAKE_EXPORT_COMPILE_COMMANDS}" STREQUAL "1")
  set(CMAKE_EXPORT_COMPILE_COMMANDS 1)
 endif()
 
-if(UNIX)
+# On Emscripten, the equivalent $ORIGIN runtime path is added
+# through the linker flags in SetupCxxFlags.cmake instead
+if(UNIX AND NOT CMAKE_SYSTEM_NAME STREQUAL "Emscripten")
  set(CMAKE_BUILD_WITH_INSTALL_RPATH TRUE)
  # In the event that we are bundling the shared libraries (e.g. in a
  # manylinux1 wheel), we need to set the RPATH of the extensions to the

python/scripts/run_emscripten_tests.py

@@ -148,8 +148,14 @@ class NodeDriver:
  import subprocess
 
  def __init__(self, hostname, port):
+ # handle different syntax for util-linux script and macOS BSD-style script
+ if sys.platform == "darwin":
+ script_command = [shutil.which("script"), "-q", "/dev/null",
+ shutil.which("node")]
+ else:
+ script_command = [shutil.which("script"), "-c", shutil.which("node")]
  self.process = subprocess.Popen(
-[shutil.which("script"), "-c", shutil.which("node")],
+script_command,
  stdin=subprocess.PIPE,
  shell=False,
  bufsize=0,
@@ -320,6 +326,7 @@ def _load_pyarrow_in_runner(driver, wheel_name):
 if "pyarrow" not in sys.modules:
  await micropip.install("hypothesis")
  import pyodide_js as pjs
+ await pjs.loadPackage("packaging")
  await pjs.loadPackage("tzdata")
  await pjs.loadPackage("numpy")
  await pjs.loadPackage("pandas")

ci/docker/conda-python-emscripten.dockerfile

@@ -18,32 +18,39 @@
 ARG repo
 ARG arch
 ARG arch_short
-ARG python="3.12"
+ARG python="3.14"
 FROM --platform=linux/${arch} ${repo}:${arch_short}-conda-python-${python}
 
 ARG selenium_version="4.41.0"
-ARG pyodide_version="0.26.0"
+ARG pyodide_version="314.0.4"
+# pyodide-build is versioned independently of Pyodide
+ARG pyodide_build_version="0.39.0"
 ARG chrome_version="latest"
-ARG required_python_min="(3,12)"
-# fail if python version < 3.12
+ARG required_python_min="(3,14)"
+# fail if python version < 3.14
 RUN echo "check PYTHON>=${required_python_min}" && python -c "import sys;sys.exit(0 if sys.version_info>=${required_python_min} else 1)"
 
-# install selenium and recent pyodide-build and recent python
+RUN apt-get update -y -q && \
+ apt-get install -y -q --no-install-recommends \
+ libatomic1 && \
+ apt-get clean && \
+ rm -rf /var/lib/apt/lists/*
 
 # needs to be a login shell so ~/.profile is read
 SHELL ["/bin/bash", "--login", "-c", "-o", "pipefail"]
 
+# install selenium and recent pyodide-build and recent python
 RUN python -m pip install --no-cache-dir selenium==${selenium_version} && \
- python -m pip install --no-cache-dir --upgrade pyodide-build>=${pyodide_version}
+ python -m pip install --no-cache-dir --upgrade "pyodide-build>=${pyodide_build_version}"
 
 # install pyodide dist directory to /pyodide
 RUN pyodide_dist_url="https://github.com/pyodide/pyodide/releases/download/${pyodide_version}/pyodide-${pyodide_version}.tar.bz2" && \
  wget -q "${pyodide_dist_url}" -O- | tar -xj -C /
 
-# install node 20 (needed for async call support)
+# install node 24 (needed for async call support and JSPI)
 # and pthread-stubs for build, and unzip needed for chrome build to work
 # xz is needed by emsdk to extract node tarballs
-RUN conda install nodejs=20 unzip pthread-stubs make xz -c conda-forge
+RUN conda install nodejs=24 unzip pthread-stubs make xz -c conda-forge
 
 # install correct version of emscripten for this pyodide
 COPY ci/scripts/install_emscripten.sh /arrow/ci/scripts/

compose.yaml

@@ -981,10 +981,11 @@ services:
  arch_short: ${ARCH_SHORT}
  clang_tools: ${CLANG_TOOLS}
  llvm: ${LLVM}
- pyodide_version: "0.26.0"
+ pyodide_version: "314.0.4"
+ pyodide_build_version: "0.39.0"
  chrome_version: "latest"
  selenium_version: "4.41.0"
- required_python_min: "(3,12)"
+ required_python_min: "(3,14)"
  python: ${PYTHON}
  shm_size: *shm-size
  volumes: *ubuntu-volumes

cpp/cmake_modules/SetupCxxFlags.cmake

@@ -719,26 +719,35 @@ if(MSVC)
 endif()
 
 if(CMAKE_SYSTEM_NAME STREQUAL "Emscripten")
- # flags are:
+ # For the Pyodide 2026_0 ABI, the flags are:
  # 1) We force *everything* to build as position independent
- # 2) And with support for C++ exceptions
- set(CMAKE_C_FLAGS "${CMAKE_C_FLAGS} -fPIC -fexceptions")
+ # 2) We support C++ WASM exceptions, using the native WebAssembly
+ # exception handling proposal and WebAssembly-based setjmp/longjmp
+ set(CMAKE_C_FLAGS "${CMAKE_C_FLAGS} -fPIC -fwasm-exceptions -sSUPPORT_LONGJMP=wasm")
  # deprecated-literal-operator error is thrown in datetime (vendored lib in arrow)
  set(CMAKE_CXX_FLAGS
- "${CMAKE_CXX_FLAGS} -fPIC -fexceptions -Wno-error=deprecated-literal-operator")
+ "${CMAKE_CXX_FLAGS} -fPIC -fwasm-exceptions -sSUPPORT_LONGJMP=wasm -Wno-error=deprecated-literal-operator"
+ )
 
  # flags for creating shared libraries (only used in pyarrow, because
  # Emscripten builds libarrow as static)
  # flags are:
  # 1) Tell it to use JavaScript / WebAssembly 64 bit number support.
- # 2) Tell it to build with support for C++ exceptions
+ # 2) Tell it to build with support for C++ exceptions (see above)
  # 3) Skip linker flags error which happens with -soname parameter
- set(ARROW_EMSCRIPTEN_LINKER_FLAGS "-sWASM_BIGINT=1 -fexceptions -Wno-error=linkflags")
- set(CMAKE_SHARED_LIBRARY_CREATE_C_FLAGS
- "-sSIDE_MODULE=1 ${ARROW_EMSCRIPTEN_LINKER_FLAGS}")
- set(CMAKE_SHARED_LIBRARY_CREATE_CXX_FLAGS
- "-sSIDE_MODULE=1 ${ARROW_EMSCRIPTEN_LINKER_FLAGS}")
- set(CMAKE_SHARED_LINKER_FLAGS "-sSIDE_MODULE=1 ${ARROW_EMSCRIPTEN_LINKER_FLAGS}")
+ # 4) Record $ORIGIN as an RPATH, so that Pyodide can locate
+ # libarrow_python.so next to the extension modules needing it
+ # Note: the CMAKE_SHARED_LIBRARY_CREATE_<LANG>_FLAGS and
+ # CMAKE_SHARED_MODULE_CREATE_<LANG>_FLAGS variables are not used by the
+ # Ninja generator, so the flags must go through the linker-flags
+ # variables, which apply to SHARED (libarrow_python) and MODULE (Cython
+ # extension) targets respectively.
+ set(ARROW_EMSCRIPTEN_LINKER_FLAGS
+ "-sWASM_BIGINT=1 -fwasm-exceptions -sSUPPORT_LONGJMP=wasm -Wno-error=linkflags")
+ set(ARROW_EMSCRIPTEN_DYLIB_CREATE_FLAGS
+ "-sSIDE_MODULE=1 ${ARROW_EMSCRIPTEN_LINKER_FLAGS} '-Wl,-rpath,$ORIGIN'")
+ set(CMAKE_SHARED_LINKER_FLAGS "${ARROW_EMSCRIPTEN_DYLIB_CREATE_FLAGS}")
+ set(CMAKE_MODULE_LINKER_FLAGS "${ARROW_EMSCRIPTEN_DYLIB_CREATE_FLAGS}")
  if(ARROW_TESTING)
  # flags for building test executables for use in node
  if("${UPPERCASE_BUILD_TYPE}" STREQUAL "RELEASE")

cpp/cmake_modules/ThirdpartyToolchain.cmake

@@ -2183,6 +2183,10 @@ function(build_protobuf)
  set(PROTOBUF_HOST_CMAKE_ARGS
  "-DCMAKE_CXX_FLAGS="
  "-DCMAKE_C_FLAGS="
+ # Explicitly set the C++ standard for the host build. Without this,
+ # Abseil may be compiled with the host compiler's default standard,
+ # which can be older than C++17.
+ "-DCMAKE_CXX_STANDARD=${CMAKE_CXX_STANDARD}"
  "-DCMAKE_INSTALL_PREFIX=${PROTOBUF_HOST_PREFIX}"
  -Dprotobuf_BUILD_TESTS=OFF
  -Dprotobuf_DEBUG_POSTFIX=

dev/tasks/tasks.yml

@@ -521,7 +521,7 @@ tasks:
  params:
  env:
  UBUNTU: 22.04
- PYTHON: 3.12
+ PYTHON: 3.14
  image: conda-python-emscripten
 
  test-conda-python-3.11-hypothesis:

docs/source/developers/cpp/emscripten.rst

@@ -34,9 +34,9 @@ activate it using the commands below (see https://emscripten.org/docs/getting_st
  git clone https://github.com/emscripten-core/emsdk.git
  cd emsdk
  # replace <version> with the desired EMSDK version.
- # e.g. for Pyodide 0.26, you need EMSDK version 3.1.58
+ # e.g. for Pyodide 314.0.x, you need EMSDK version 5.0.3
  # the versions can be found in the Makefile.envs file in the Pyodide repo:
- # https://github.com/pyodide/pyodide/blob/10b484cfe427e076c929a55dc35cfff01ea8d3bc/Makefile.envs
+ # https://github.com/pyodide/pyodide/blob/314.0.4/Makefile.envs
  ./emsdk install <version>
  ./emsdk activate <version>
  source ./emsdk_env.sh

python/CMakeLists.txt

@@ -111,7 +111,9 @@ if("$ENV{CMAKE_EXPORT_COMPILE_COMMANDS}" STREQUAL "1")
  set(CMAKE_EXPORT_COMPILE_COMMANDS 1)
 endif()
 
-if(UNIX)
+# On Emscripten, the equivalent $ORIGIN runtime path is added
+# through the linker flags in SetupCxxFlags.cmake instead
+if(UNIX AND NOT CMAKE_SYSTEM_NAME STREQUAL "Emscripten")
  set(CMAKE_BUILD_WITH_INSTALL_RPATH TRUE)
  # In the event that we are bundling the shared libraries (e.g. in a
  # manylinux1 wheel), we need to set the RPATH of the extensions to the

python/scripts/run_emscripten_tests.py

@@ -148,8 +148,14 @@ class NodeDriver:
  import subprocess
 
  def __init__(self, hostname, port):
+ # handle different syntax for util-linux script and macOS BSD-style script
+ if sys.platform == "darwin":
+ script_command = [shutil.which("script"), "-q", "/dev/null",
+ shutil.which("node")]
+ else:
+ script_command = [shutil.which("script"), "-c", shutil.which("node")]
  self.process = subprocess.Popen(
-[shutil.which("script"), "-c", shutil.which("node")],
+script_command,
  stdin=subprocess.PIPE,
  shell=False,
  bufsize=0,
@@ -320,6 +326,7 @@ def _load_pyarrow_in_runner(driver, wheel_name):
 if "pyarrow" not in sys.modules:
  await micropip.install("hypothesis")
  import pyodide_js as pjs
+ await pjs.loadPackage("packaging")
  await pjs.loadPackage("tzdata")
  await pjs.loadPackage("numpy")
  await pjs.loadPackage("pandas")

ci/docker/conda-python-emscripten.dockerfile

@@ -18,32 +18,39 @@
 ARG repo
 ARG arch
 ARG arch_short
-ARG python="3.12"
+ARG python="3.14"
 FROM --platform=linux/${arch} ${repo}:${arch_short}-conda-python-${python}
 
 ARG selenium_version="4.41.0"
-ARG pyodide_version="0.26.0"
+ARG pyodide_version="314.0.4"
+# pyodide-build is versioned independently of Pyodide
+ARG pyodide_build_version="0.39.0"
 ARG chrome_version="latest"
-ARG required_python_min="(3,12)"
-# fail if python version < 3.12
+ARG required_python_min="(3,14)"
+# fail if python version < 3.14
 RUN echo "check PYTHON>=${required_python_min}" && python -c "import sys;sys.exit(0 if sys.version_info>=${required_python_min} else 1)"
 
-# install selenium and recent pyodide-build and recent python
+RUN apt-get update -y -q && \
+ apt-get install -y -q --no-install-recommends \
+ libatomic1 && \
+ apt-get clean && \
+ rm -rf /var/lib/apt/lists/*
 
 # needs to be a login shell so ~/.profile is read
 SHELL ["/bin/bash", "--login", "-c", "-o", "pipefail"]
 
+# install selenium and recent pyodide-build and recent python
 RUN python -m pip install --no-cache-dir selenium==${selenium_version} && \
- python -m pip install --no-cache-dir --upgrade pyodide-build>=${pyodide_version}
+ python -m pip install --no-cache-dir --upgrade "pyodide-build>=${pyodide_build_version}"
 
 # install pyodide dist directory to /pyodide
 RUN pyodide_dist_url="https://github.com/pyodide/pyodide/releases/download/${pyodide_version}/pyodide-${pyodide_version}.tar.bz2" && \
  wget -q "${pyodide_dist_url}" -O- | tar -xj -C /
 
-# install node 20 (needed for async call support)
+# install node 24 (needed for async call support and JSPI)
 # and pthread-stubs for build, and unzip needed for chrome build to work
 # xz is needed by emsdk to extract node tarballs
-RUN conda install nodejs=20 unzip pthread-stubs make xz -c conda-forge
+RUN conda install nodejs=24 unzip pthread-stubs make xz -c conda-forge
 
 # install correct version of emscripten for this pyodide
 COPY ci/scripts/install_emscripten.sh /arrow/ci/scripts/

compose.yaml

@@ -981,10 +981,11 @@ services:
  arch_short: ${ARCH_SHORT}
  clang_tools: ${CLANG_TOOLS}
  llvm: ${LLVM}
- pyodide_version: "0.26.0"
+ pyodide_version: "314.0.4"
+ pyodide_build_version: "0.39.0"
  chrome_version: "latest"
  selenium_version: "4.41.0"
- required_python_min: "(3,12)"
+ required_python_min: "(3,14)"
  python: ${PYTHON}
  shm_size: *shm-size
  volumes: *ubuntu-volumes

cpp/cmake_modules/SetupCxxFlags.cmake

@@ -719,26 +719,35 @@ if(MSVC)
 endif()
 
 if(CMAKE_SYSTEM_NAME STREQUAL "Emscripten")
- # flags are:
+ # For the Pyodide 2026_0 ABI, the flags are:
  # 1) We force *everything* to build as position independent
- # 2) And with support for C++ exceptions
- set(CMAKE_C_FLAGS "${CMAKE_C_FLAGS} -fPIC -fexceptions")
+ # 2) We support C++ WASM exceptions, using the native WebAssembly
+ # exception handling proposal and WebAssembly-based setjmp/longjmp
+ set(CMAKE_C_FLAGS "${CMAKE_C_FLAGS} -fPIC -fwasm-exceptions -sSUPPORT_LONGJMP=wasm")
  # deprecated-literal-operator error is thrown in datetime (vendored lib in arrow)
  set(CMAKE_CXX_FLAGS
- "${CMAKE_CXX_FLAGS} -fPIC -fexceptions -Wno-error=deprecated-literal-operator")
+ "${CMAKE_CXX_FLAGS} -fPIC -fwasm-exceptions -sSUPPORT_LONGJMP=wasm -Wno-error=deprecated-literal-operator"
+ )
 
  # flags for creating shared libraries (only used in pyarrow, because
  # Emscripten builds libarrow as static)
  # flags are:
  # 1) Tell it to use JavaScript / WebAssembly 64 bit number support.
- # 2) Tell it to build with support for C++ exceptions
+ # 2) Tell it to build with support for C++ exceptions (see above)
  # 3) Skip linker flags error which happens with -soname parameter
- set(ARROW_EMSCRIPTEN_LINKER_FLAGS "-sWASM_BIGINT=1 -fexceptions -Wno-error=linkflags")
- set(CMAKE_SHARED_LIBRARY_CREATE_C_FLAGS
- "-sSIDE_MODULE=1 ${ARROW_EMSCRIPTEN_LINKER_FLAGS}")
- set(CMAKE_SHARED_LIBRARY_CREATE_CXX_FLAGS
- "-sSIDE_MODULE=1 ${ARROW_EMSCRIPTEN_LINKER_FLAGS}")
- set(CMAKE_SHARED_LINKER_FLAGS "-sSIDE_MODULE=1 ${ARROW_EMSCRIPTEN_LINKER_FLAGS}")
+ # 4) Record $ORIGIN as an RPATH, so that Pyodide can locate
+ # libarrow_python.so next to the extension modules needing it
+ # Note: the CMAKE_SHARED_LIBRARY_CREATE_<LANG>_FLAGS and
+ # CMAKE_SHARED_MODULE_CREATE_<LANG>_FLAGS variables are not used by the
+ # Ninja generator, so the flags must go through the linker-flags
+ # variables, which apply to SHARED (libarrow_python) and MODULE (Cython
+ # extension) targets respectively.
+ set(ARROW_EMSCRIPTEN_LINKER_FLAGS
+ "-sWASM_BIGINT=1 -fwasm-exceptions -sSUPPORT_LONGJMP=wasm -Wno-error=linkflags")
+ set(ARROW_EMSCRIPTEN_DYLIB_CREATE_FLAGS
+ "-sSIDE_MODULE=1 ${ARROW_EMSCRIPTEN_LINKER_FLAGS} '-Wl,-rpath,$ORIGIN'")
+ set(CMAKE_SHARED_LINKER_FLAGS "${ARROW_EMSCRIPTEN_DYLIB_CREATE_FLAGS}")
+ set(CMAKE_MODULE_LINKER_FLAGS "${ARROW_EMSCRIPTEN_DYLIB_CREATE_FLAGS}")
  if(ARROW_TESTING)
  # flags for building test executables for use in node
  if("${UPPERCASE_BUILD_TYPE}" STREQUAL "RELEASE")

cpp/cmake_modules/ThirdpartyToolchain.cmake

@@ -2183,6 +2183,10 @@ function(build_protobuf)
  set(PROTOBUF_HOST_CMAKE_ARGS
  "-DCMAKE_CXX_FLAGS="
  "-DCMAKE_C_FLAGS="
+ # Explicitly set the C++ standard for the host build. Without this,
+ # Abseil may be compiled with the host compiler's default standard,
+ # which can be older than C++17.
+ "-DCMAKE_CXX_STANDARD=${CMAKE_CXX_STANDARD}"
  "-DCMAKE_INSTALL_PREFIX=${PROTOBUF_HOST_PREFIX}"
  -Dprotobuf_BUILD_TESTS=OFF
  -Dprotobuf_DEBUG_POSTFIX=

dev/tasks/tasks.yml

@@ -521,7 +521,7 @@ tasks:
  params:
  env:
  UBUNTU: 22.04
- PYTHON: 3.12
+ PYTHON: 3.14
  image: conda-python-emscripten
 
  test-conda-python-3.11-hypothesis:

docs/source/developers/cpp/emscripten.rst

@@ -34,9 +34,9 @@ activate it using the commands below (see https://emscripten.org/docs/getting_st
  git clone https://github.com/emscripten-core/emsdk.git
  cd emsdk
  # replace <version> with the desired EMSDK version.
- # e.g. for Pyodide 0.26, you need EMSDK version 3.1.58
+ # e.g. for Pyodide 314.0.x, you need EMSDK version 5.0.3
  # the versions can be found in the Makefile.envs file in the Pyodide repo:
- # https://github.com/pyodide/pyodide/blob/10b484cfe427e076c929a55dc35cfff01ea8d3bc/Makefile.envs
+ # https://github.com/pyodide/pyodide/blob/314.0.4/Makefile.envs
  ./emsdk install <version>
  ./emsdk activate <version>
  source ./emsdk_env.sh

python/CMakeLists.txt

@@ -111,7 +111,9 @@ if("$ENV{CMAKE_EXPORT_COMPILE_COMMANDS}" STREQUAL "1")
  set(CMAKE_EXPORT_COMPILE_COMMANDS 1)
 endif()
 
-if(UNIX)
+# On Emscripten, the equivalent $ORIGIN runtime path is added
+# through the linker flags in SetupCxxFlags.cmake instead
+if(UNIX AND NOT CMAKE_SYSTEM_NAME STREQUAL "Emscripten")
  set(CMAKE_BUILD_WITH_INSTALL_RPATH TRUE)
  # In the event that we are bundling the shared libraries (e.g. in a
  # manylinux1 wheel), we need to set the RPATH of the extensions to the

python/scripts/run_emscripten_tests.py

@@ -148,8 +148,14 @@ class NodeDriver:
  import subprocess
 
  def __init__(self, hostname, port):
+ # handle different syntax for util-linux script and macOS BSD-style script
+ if sys.platform == "darwin":
+ script_command = [shutil.which("script"), "-q", "/dev/null",
+ shutil.which("node")]
+ else:
+ script_command = [shutil.which("script"), "-c", shutil.which("node")]
  self.process = subprocess.Popen(
-[shutil.which("script"), "-c", shutil.which("node")],
+script_command,
  stdin=subprocess.PIPE,
  shell=False,
  bufsize=0,
@@ -320,6 +326,7 @@ def _load_pyarrow_in_runner(driver, wheel_name):
 if "pyarrow" not in sys.modules:
  await micropip.install("hypothesis")
  import pyodide_js as pjs
+ await pjs.loadPackage("packaging")
  await pjs.loadPackage("tzdata")
  await pjs.loadPackage("numpy")
  await pjs.loadPackage("pandas")

ci/docker/conda-python-emscripten.dockerfile

@@ -18,32 +18,39 @@
 ARG repo
 ARG arch
 ARG arch_short
-ARG python="3.12"
+ARG python="3.14"
 FROM --platform=linux/${arch} ${repo}:${arch_short}-conda-python-${python}
 
 ARG selenium_version="4.41.0"
-ARG pyodide_version="0.26.0"
+ARG pyodide_version="314.0.4"
+# pyodide-build is versioned independently of Pyodide
+ARG pyodide_build_version="0.39.0"
 ARG chrome_version="latest"
-ARG required_python_min="(3,12)"
-# fail if python version < 3.12
+ARG required_python_min="(3,14)"
+# fail if python version < 3.14
 RUN echo "check PYTHON>=${required_python_min}" && python -c "import sys;sys.exit(0 if sys.version_info>=${required_python_min} else 1)"
 
-# install selenium and recent pyodide-build and recent python
+RUN apt-get update -y -q && \
+ apt-get install -y -q --no-install-recommends \
+ libatomic1 && \
+ apt-get clean && \
+ rm -rf /var/lib/apt/lists/*
 
 # needs to be a login shell so ~/.profile is read
 SHELL ["/bin/bash", "--login", "-c", "-o", "pipefail"]
 
+# install selenium and recent pyodide-build and recent python
 RUN python -m pip install --no-cache-dir selenium==${selenium_version} && \
- python -m pip install --no-cache-dir --upgrade pyodide-build>=${pyodide_version}
+ python -m pip install --no-cache-dir --upgrade "pyodide-build>=${pyodide_build_version}"
 
 # install pyodide dist directory to /pyodide
 RUN pyodide_dist_url="https://github.com/pyodide/pyodide/releases/download/${pyodide_version}/pyodide-${pyodide_version}.tar.bz2" && \
  wget -q "${pyodide_dist_url}" -O- | tar -xj -C /
 
-# install node 20 (needed for async call support)
+# install node 24 (needed for async call support and JSPI)
 # and pthread-stubs for build, and unzip needed for chrome build to work
 # xz is needed by emsdk to extract node tarballs
-RUN conda install nodejs=20 unzip pthread-stubs make xz -c conda-forge
+RUN conda install nodejs=24 unzip pthread-stubs make xz -c conda-forge
 
 # install correct version of emscripten for this pyodide
 COPY ci/scripts/install_emscripten.sh /arrow/ci/scripts/

compose.yaml

@@ -981,10 +981,11 @@ services:
  arch_short: ${ARCH_SHORT}
  clang_tools: ${CLANG_TOOLS}
  llvm: ${LLVM}
- pyodide_version: "0.26.0"
+ pyodide_version: "314.0.4"
+ pyodide_build_version: "0.39.0"
  chrome_version: "latest"
  selenium_version: "4.41.0"
- required_python_min: "(3,12)"
+ required_python_min: "(3,14)"
  python: ${PYTHON}
  shm_size: *shm-size
  volumes: *ubuntu-volumes

cpp/cmake_modules/SetupCxxFlags.cmake

@@ -719,26 +719,35 @@ if(MSVC)
 endif()
 
 if(CMAKE_SYSTEM_NAME STREQUAL "Emscripten")
- # flags are:
+ # For the Pyodide 2026_0 ABI, the flags are:
  # 1) We force *everything* to build as position independent
- # 2) And with support for C++ exceptions
- set(CMAKE_C_FLAGS "${CMAKE_C_FLAGS} -fPIC -fexceptions")
+ # 2) We support C++ WASM exceptions, using the native WebAssembly
+ # exception handling proposal and WebAssembly-based setjmp/longjmp
+ set(CMAKE_C_FLAGS "${CMAKE_C_FLAGS} -fPIC -fwasm-exceptions -sSUPPORT_LONGJMP=wasm")
  # deprecated-literal-operator error is thrown in datetime (vendored lib in arrow)
  set(CMAKE_CXX_FLAGS
- "${CMAKE_CXX_FLAGS} -fPIC -fexceptions -Wno-error=deprecated-literal-operator")
+ "${CMAKE_CXX_FLAGS} -fPIC -fwasm-exceptions -sSUPPORT_LONGJMP=wasm -Wno-error=deprecated-literal-operator"
+ )
 
  # flags for creating shared libraries (only used in pyarrow, because
  # Emscripten builds libarrow as static)
  # flags are:
  # 1) Tell it to use JavaScript / WebAssembly 64 bit number support.
- # 2) Tell it to build with support for C++ exceptions
+ # 2) Tell it to build with support for C++ exceptions (see above)
  # 3) Skip linker flags error which happens with -soname parameter
- set(ARROW_EMSCRIPTEN_LINKER_FLAGS "-sWASM_BIGINT=1 -fexceptions -Wno-error=linkflags")
- set(CMAKE_SHARED_LIBRARY_CREATE_C_FLAGS
- "-sSIDE_MODULE=1 ${ARROW_EMSCRIPTEN_LINKER_FLAGS}")
- set(CMAKE_SHARED_LIBRARY_CREATE_CXX_FLAGS
- "-sSIDE_MODULE=1 ${ARROW_EMSCRIPTEN_LINKER_FLAGS}")
- set(CMAKE_SHARED_LINKER_FLAGS "-sSIDE_MODULE=1 ${ARROW_EMSCRIPTEN_LINKER_FLAGS}")
+ # 4) Record $ORIGIN as an RPATH, so that Pyodide can locate
+ # libarrow_python.so next to the extension modules needing it
+ # Note: the CMAKE_SHARED_LIBRARY_CREATE_<LANG>_FLAGS and
+ # CMAKE_SHARED_MODULE_CREATE_<LANG>_FLAGS variables are not used by the
+ # Ninja generator, so the flags must go through the linker-flags
+ # variables, which apply to SHARED (libarrow_python) and MODULE (Cython
+ # extension) targets respectively.
+ set(ARROW_EMSCRIPTEN_LINKER_FLAGS
+ "-sWASM_BIGINT=1 -fwasm-exceptions -sSUPPORT_LONGJMP=wasm -Wno-error=linkflags")
+ set(ARROW_EMSCRIPTEN_DYLIB_CREATE_FLAGS
+ "-sSIDE_MODULE=1 ${ARROW_EMSCRIPTEN_LINKER_FLAGS} '-Wl,-rpath,$ORIGIN'")
+ set(CMAKE_SHARED_LINKER_FLAGS "${ARROW_EMSCRIPTEN_DYLIB_CREATE_FLAGS}")
+ set(CMAKE_MODULE_LINKER_FLAGS "${ARROW_EMSCRIPTEN_DYLIB_CREATE_FLAGS}")
  if(ARROW_TESTING)
  # flags for building test executables for use in node
  if("${UPPERCASE_BUILD_TYPE}" STREQUAL "RELEASE")

cpp/cmake_modules/ThirdpartyToolchain.cmake

@@ -2183,6 +2183,10 @@ function(build_protobuf)
  set(PROTOBUF_HOST_CMAKE_ARGS
  "-DCMAKE_CXX_FLAGS="
  "-DCMAKE_C_FLAGS="
+ # Explicitly set the C++ standard for the host build. Without this,
+ # Abseil may be compiled with the host compiler's default standard,
+ # which can be older than C++17.
+ "-DCMAKE_CXX_STANDARD=${CMAKE_CXX_STANDARD}"
  "-DCMAKE_INSTALL_PREFIX=${PROTOBUF_HOST_PREFIX}"
  -Dprotobuf_BUILD_TESTS=OFF
  -Dprotobuf_DEBUG_POSTFIX=

dev/tasks/tasks.yml

@@ -521,7 +521,7 @@ tasks:
  params:
  env:
  UBUNTU: 22.04
- PYTHON: 3.12
+ PYTHON: 3.14
  image: conda-python-emscripten
 
  test-conda-python-3.11-hypothesis:

docs/source/developers/cpp/emscripten.rst

@@ -34,9 +34,9 @@ activate it using the commands below (see https://emscripten.org/docs/getting_st
  git clone https://github.com/emscripten-core/emsdk.git
  cd emsdk
  # replace <version> with the desired EMSDK version.
- # e.g. for Pyodide 0.26, you need EMSDK version 3.1.58
+ # e.g. for Pyodide 314.0.x, you need EMSDK version 5.0.3
  # the versions can be found in the Makefile.envs file in the Pyodide repo:
- # https://github.com/pyodide/pyodide/blob/10b484cfe427e076c929a55dc35cfff01ea8d3bc/Makefile.envs
+ # https://github.com/pyodide/pyodide/blob/314.0.4/Makefile.envs
  ./emsdk install <version>
  ./emsdk activate <version>
  source ./emsdk_env.sh

python/CMakeLists.txt

@@ -111,7 +111,9 @@ if("$ENV{CMAKE_EXPORT_COMPILE_COMMANDS}" STREQUAL "1")
  set(CMAKE_EXPORT_COMPILE_COMMANDS 1)
 endif()
 
-if(UNIX)
+# On Emscripten, the equivalent $ORIGIN runtime path is added
+# through the linker flags in SetupCxxFlags.cmake instead
+if(UNIX AND NOT CMAKE_SYSTEM_NAME STREQUAL "Emscripten")
  set(CMAKE_BUILD_WITH_INSTALL_RPATH TRUE)
  # In the event that we are bundling the shared libraries (e.g. in a
  # manylinux1 wheel), we need to set the RPATH of the extensions to the

python/scripts/run_emscripten_tests.py

@@ -148,8 +148,14 @@ class NodeDriver:
  import subprocess
 
  def __init__(self, hostname, port):
+ # handle different syntax for util-linux script and macOS BSD-style script
+ if sys.platform == "darwin":
+ script_command = [shutil.which("script"), "-q", "/dev/null",
+ shutil.which("node")]
+ else:
+ script_command = [shutil.which("script"), "-c", shutil.which("node")]
  self.process = subprocess.Popen(
-[shutil.which("script"), "-c", shutil.which("node")],
+script_command,
  stdin=subprocess.PIPE,
  shell=False,
  bufsize=0,
@@ -320,6 +326,7 @@ def _load_pyarrow_in_runner(driver, wheel_name):
 if "pyarrow" not in sys.modules:
  await micropip.install("hypothesis")
  import pyodide_js as pjs
+ await pjs.loadPackage("packaging")
  await pjs.loadPackage("tzdata")
  await pjs.loadPackage("numpy")
  await pjs.loadPackage("pandas")

ci/docker/conda-python-emscripten.dockerfile

@@ -18,32 +18,39 @@
 ARG repo
 ARG arch
 ARG arch_short
-ARG python="3.12"
+ARG python="3.14"
 FROM --platform=linux/${arch} ${repo}:${arch_short}-conda-python-${python}
 
 ARG selenium_version="4.41.0"
-ARG pyodide_version="0.26.0"
+ARG pyodide_version="314.0.4"
+# pyodide-build is versioned independently of Pyodide
+ARG pyodide_build_version="0.39.0"
 ARG chrome_version="latest"
-ARG required_python_min="(3,12)"
-# fail if python version < 3.12
+ARG required_python_min="(3,14)"
+# fail if python version < 3.14
 RUN echo "check PYTHON>=${required_python_min}" && python -c "import sys;sys.exit(0 if sys.version_info>=${required_python_min} else 1)"
 
-# install selenium and recent pyodide-build and recent python
+RUN apt-get update -y -q && \
+ apt-get install -y -q --no-install-recommends \
+ libatomic1 && \
+ apt-get clean && \
+ rm -rf /var/lib/apt/lists/*
 
 # needs to be a login shell so ~/.profile is read
 SHELL ["/bin/bash", "--login", "-c", "-o", "pipefail"]
 
+# install selenium and recent pyodide-build and recent python
 RUN python -m pip install --no-cache-dir selenium==${selenium_version} && \
- python -m pip install --no-cache-dir --upgrade pyodide-build>=${pyodide_version}
+ python -m pip install --no-cache-dir --upgrade "pyodide-build>=${pyodide_build_version}"
 
 # install pyodide dist directory to /pyodide
 RUN pyodide_dist_url="https://github.com/pyodide/pyodide/releases/download/${pyodide_version}/pyodide-${pyodide_version}.tar.bz2" && \
  wget -q "${pyodide_dist_url}" -O- | tar -xj -C /
 
-# install node 20 (needed for async call support)
+# install node 24 (needed for async call support and JSPI)
 # and pthread-stubs for build, and unzip needed for chrome build to work
 # xz is needed by emsdk to extract node tarballs
-RUN conda install nodejs=20 unzip pthread-stubs make xz -c conda-forge
+RUN conda install nodejs=24 unzip pthread-stubs make xz -c conda-forge
 
 # install correct version of emscripten for this pyodide
 COPY ci/scripts/install_emscripten.sh /arrow/ci/scripts/

compose.yaml

@@ -981,10 +981,11 @@ services:
  arch_short: ${ARCH_SHORT}
  clang_tools: ${CLANG_TOOLS}
  llvm: ${LLVM}
- pyodide_version: "0.26.0"
+ pyodide_version: "314.0.4"
+ pyodide_build_version: "0.39.0"
  chrome_version: "latest"
  selenium_version: "4.41.0"
- required_python_min: "(3,12)"
+ required_python_min: "(3,14)"
  python: ${PYTHON}
  shm_size: *shm-size
  volumes: *ubuntu-volumes

cpp/cmake_modules/SetupCxxFlags.cmake

@@ -719,26 +719,35 @@ if(MSVC)
 endif()
 
 if(CMAKE_SYSTEM_NAME STREQUAL "Emscripten")
- # flags are:
+ # For the Pyodide 2026_0 ABI, the flags are:
  # 1) We force *everything* to build as position independent
- # 2) And with support for C++ exceptions
- set(CMAKE_C_FLAGS "${CMAKE_C_FLAGS} -fPIC -fexceptions")
+ # 2) We support C++ WASM exceptions, using the native WebAssembly
+ # exception handling proposal and WebAssembly-based setjmp/longjmp
+ set(CMAKE_C_FLAGS "${CMAKE_C_FLAGS} -fPIC -fwasm-exceptions -sSUPPORT_LONGJMP=wasm")
  # deprecated-literal-operator error is thrown in datetime (vendored lib in arrow)
  set(CMAKE_CXX_FLAGS
- "${CMAKE_CXX_FLAGS} -fPIC -fexceptions -Wno-error=deprecated-literal-operator")
+ "${CMAKE_CXX_FLAGS} -fPIC -fwasm-exceptions -sSUPPORT_LONGJMP=wasm -Wno-error=deprecated-literal-operator"
+ )
 
  # flags for creating shared libraries (only used in pyarrow, because
  # Emscripten builds libarrow as static)
  # flags are:
  # 1) Tell it to use JavaScript / WebAssembly 64 bit number support.
- # 2) Tell it to build with support for C++ exceptions
+ # 2) Tell it to build with support for C++ exceptions (see above)
  # 3) Skip linker flags error which happens with -soname parameter
- set(ARROW_EMSCRIPTEN_LINKER_FLAGS "-sWASM_BIGINT=1 -fexceptions -Wno-error=linkflags")
- set(CMAKE_SHARED_LIBRARY_CREATE_C_FLAGS
- "-sSIDE_MODULE=1 ${ARROW_EMSCRIPTEN_LINKER_FLAGS}")
- set(CMAKE_SHARED_LIBRARY_CREATE_CXX_FLAGS
- "-sSIDE_MODULE=1 ${ARROW_EMSCRIPTEN_LINKER_FLAGS}")
- set(CMAKE_SHARED_LINKER_FLAGS "-sSIDE_MODULE=1 ${ARROW_EMSCRIPTEN_LINKER_FLAGS}")
+ # 4) Record $ORIGIN as an RPATH, so that Pyodide can locate
+ # libarrow_python.so next to the extension modules needing it
+ # Note: the CMAKE_SHARED_LIBRARY_CREATE_<LANG>_FLAGS and
+ # CMAKE_SHARED_MODULE_CREATE_<LANG>_FLAGS variables are not used by the
+ # Ninja generator, so the flags must go through the linker-flags
+ # variables, which apply to SHARED (libarrow_python) and MODULE (Cython
+ # extension) targets respectively.
+ set(ARROW_EMSCRIPTEN_LINKER_FLAGS
+ "-sWASM_BIGINT=1 -fwasm-exceptions -sSUPPORT_LONGJMP=wasm -Wno-error=linkflags")
+ set(ARROW_EMSCRIPTEN_DYLIB_CREATE_FLAGS
+ "-sSIDE_MODULE=1 ${ARROW_EMSCRIPTEN_LINKER_FLAGS} '-Wl,-rpath,$ORIGIN'")
+ set(CMAKE_SHARED_LINKER_FLAGS "${ARROW_EMSCRIPTEN_DYLIB_CREATE_FLAGS}")
+ set(CMAKE_MODULE_LINKER_FLAGS "${ARROW_EMSCRIPTEN_DYLIB_CREATE_FLAGS}")
  if(ARROW_TESTING)
  # flags for building test executables for use in node
  if("${UPPERCASE_BUILD_TYPE}" STREQUAL "RELEASE")

cpp/cmake_modules/ThirdpartyToolchain.cmake

@@ -2183,6 +2183,10 @@ function(build_protobuf)
  set(PROTOBUF_HOST_CMAKE_ARGS
  "-DCMAKE_CXX_FLAGS="
  "-DCMAKE_C_FLAGS="
+ # Explicitly set the C++ standard for the host build. Without this,
+ # Abseil may be compiled with the host compiler's default standard,
+ # which can be older than C++17.
+ "-DCMAKE_CXX_STANDARD=${CMAKE_CXX_STANDARD}"
  "-DCMAKE_INSTALL_PREFIX=${PROTOBUF_HOST_PREFIX}"
  -Dprotobuf_BUILD_TESTS=OFF
  -Dprotobuf_DEBUG_POSTFIX=

dev/tasks/tasks.yml

@@ -521,7 +521,7 @@ tasks:
  params:
  env:
  UBUNTU: 22.04
- PYTHON: 3.12
+ PYTHON: 3.14
  image: conda-python-emscripten
 
  test-conda-python-3.11-hypothesis:

docs/source/developers/cpp/emscripten.rst

@@ -34,9 +34,9 @@ activate it using the commands below (see https://emscripten.org/docs/getting_st
  git clone https://github.com/emscripten-core/emsdk.git
  cd emsdk
  # replace <version> with the desired EMSDK version.
- # e.g. for Pyodide 0.26, you need EMSDK version 3.1.58
+ # e.g. for Pyodide 314.0.x, you need EMSDK version 5.0.3
  # the versions can be found in the Makefile.envs file in the Pyodide repo:
- # https://github.com/pyodide/pyodide/blob/10b484cfe427e076c929a55dc35cfff01ea8d3bc/Makefile.envs
+ # https://github.com/pyodide/pyodide/blob/314.0.4/Makefile.envs
  ./emsdk install <version>
  ./emsdk activate <version>
  source ./emsdk_env.sh

python/CMakeLists.txt

@@ -111,7 +111,9 @@ if("$ENV{CMAKE_EXPORT_COMPILE_COMMANDS}" STREQUAL "1")
  set(CMAKE_EXPORT_COMPILE_COMMANDS 1)
 endif()
 
-if(UNIX)
+# On Emscripten, the equivalent $ORIGIN runtime path is added
+# through the linker flags in SetupCxxFlags.cmake instead
+if(UNIX AND NOT CMAKE_SYSTEM_NAME STREQUAL "Emscripten")
  set(CMAKE_BUILD_WITH_INSTALL_RPATH TRUE)
  # In the event that we are bundling the shared libraries (e.g. in a
  # manylinux1 wheel), we need to set the RPATH of the extensions to the

python/scripts/run_emscripten_tests.py

@@ -148,8 +148,14 @@ class NodeDriver:
  import subprocess
 
  def __init__(self, hostname, port):
+ # handle different syntax for util-linux script and macOS BSD-style script
+ if sys.platform == "darwin":
+ script_command = [shutil.which("script"), "-q", "/dev/null",
+ shutil.which("node")]
+ else:
+ script_command = [shutil.which("script"), "-c", shutil.which("node")]
  self.process = subprocess.Popen(
-[shutil.which("script"), "-c", shutil.which("node")],
+script_command,
  stdin=subprocess.PIPE,
  shell=False,
  bufsize=0,
@@ -320,6 +326,7 @@ def _load_pyarrow_in_runner(driver, wheel_name):
 if "pyarrow" not in sys.modules:
  await micropip.install("hypothesis")
  import pyodide_js as pjs
+ await pjs.loadPackage("packaging")
  await pjs.loadPackage("tzdata")
  await pjs.loadPackage("numpy")
  await pjs.loadPackage("pandas")