From 1b269658847759c04504c19b5c2ddc52eaedbeef Mon Sep 17 00:00:00 2001
From: Joe Marshall <joe.marshall@nottingham.ac.uk>
Date: Mon, 12 Aug 2024 10:10:20 +0100
Subject: [PATCH] initial github action test

---
 .github/workflows/make_wheels.yaml       | 41 ++++++++++++++++++
 .gitignore                               |  1 +
 README.md                                |  3 ++
 pyarrow/meta.yaml                        | 55 ++++++++++++++++++++++++
 pyarrow/patches/autoload_timezones.patch | 13 ++++++
 scripts/get_pyodide_versions.sh          |  6 +++
 6 files changed, 119 insertions(+)
 create mode 100644 .github/workflows/make_wheels.yaml
 create mode 100644 .gitignore
 create mode 100644 README.md
 create mode 100644 pyarrow/meta.yaml
 create mode 100644 pyarrow/patches/autoload_timezones.patch
 create mode 100644 scripts/get_pyodide_versions.sh

diff --git a/.github/workflows/make_wheels.yaml b/.github/workflows/make_wheels.yaml
new file mode 100644
index 0000000..3525359
--- /dev/null
+++ b/.github/workflows/make_wheels.yaml
@@ -0,0 +1,41 @@
+name: main
+
+on:
+  push:
+  pull_request:
+  # Allows you to run this workflow manually from the Actions tab
+  workflow_dispatch:
+
+permissions:
+  contents: read
+
+jobs:
+  build:
+    strategy:
+        matrix:
+          include:
+          - pyodide: 0.27.3
+            python: 3.12
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+      - uses: actions/setup-python@v5
+        with:
+          python-version: ${{ matrix.python }}
+      - name: install pyodide build
+        run: pip install pyodide-build=${{ matrix.pyodide }}
+      - name: Get emscripten version
+        run: python -c 'from pyodide_build.build_env import emscripten_version;print(f"EMVERSION={emscripten_version()}")' >> $GITHUB_ENV
+      - name: Install emsdk ${{env.EMVERSION}}
+        uses: mymindstorm/setup-emsdk@v14
+        with:
+          version: ${{env.EMVERSION}}
+          actions-cache-folder: 'emsdk-cache'
+      - run: mkdir dist
+      - name: Arrow build
+        run: pyodide build-recipe-no-deps --recipe-dir . --target-dir dist pyarrow
+      - uses: actions/upload-artifact@v4
+        with:
+          name: build_wheel
+          path: dist
+
diff --git a/.gitignore b/.gitignore
new file mode 100644
index 0000000..98e0579
--- /dev/null
+++ b/.gitignore
@@ -0,0 +1 @@
+.pyodide*
\ No newline at end of file
diff --git a/README.md b/README.md
new file mode 100644
index 0000000..ae5ef9f
--- /dev/null
+++ b/README.md
@@ -0,0 +1,3 @@
+# PyArrow build for pyodide
+
+This builds pyarrow wheels for use in pyodide. It is separate to the main pyodide build because pyarrow is a SLOW build.
\ No newline at end of file
diff --git a/pyarrow/meta.yaml b/pyarrow/meta.yaml
new file mode 100644
index 0000000..f3030af
--- /dev/null
+++ b/pyarrow/meta.yaml
@@ -0,0 +1,55 @@
+package:
+  name: pyarrow
+  version: 17.0.0
+  top-level:
+    - pyarrow
+source:
+  url: https://github.com/apache/arrow/releases/download/apache-arrow-17.0.0/apache-arrow-17.0.0.tar.gz
+  sha256: 9d280d8042e7cf526f8c28d170d93bfab65e50f94569f6a790982a878d8d898d
+  #  extract_dir: arrow-57ac40ca23ebcaa4d42ae808137033689d3be9b1
+  patches:
+    - patches/autoload_timezones.patch
+build:
+  script: |
+    # move things around so that the top level folder is the python folder
+    # with the pyproject.toml, so pyodide build works
+    cd ..
+    # build libarrow as static library without the pyodide toolchain file
+    # because that does bad things to the build
+    BACKUP_TOOLCHAIN=${CMAKE_TOOLCHAIN_FILE}
+    unset CMAKE_TOOLCHAIN_FILE
+    if [ ! -d "${PKG_BUILD_DIR}_cpp" ]; then
+      mv $PKG_BUILD_DIR ${PKG_BUILD_DIR}_cpp
+      cp -r ${PKG_BUILD_DIR}_cpp/python $PKG_BUILD_DIR
+      # build arrow cpp library and install to emscripten sysroot
+      # n.b. the if statements below mean that you can call pyodide build-packages-no-deps
+      # pyarrow --continue to debug build errors without having to do a full rebuild
+      mkdir cppbuild
+      cd cppbuild
+      emcmake cmake --preset ninja-release-emscripten ${PKG_BUILD_DIR}_cpp/cpp
+    else
+      cd cppbuild
+    fi
+    cmake --build . --target=protobuf_ep_host -j 4
+    cmake --build . --target=install -j 4
+    cd ${PKG_BUILD_DIR}
+    export CMAKE_TOOLCHAIN_FILE=${BACKUP_TOOLCHAIN}
+    rm -f CMakeCache.txt
+    rm -rf CMakeFiles
+  cross-script: |
+    cmake --version
+about:
+  home: https://arrow.apache.org/
+  PyPI: https://pypi.org/project/pyarrow
+  summary: Python library for Apache Arrow
+  license: Apache License, Version 2.0
+requirements:
+  run:
+    - numpy
+    - pandas
+    - pyodide-unix-timezones
+  host:
+    - numpy
+extra:
+  recipe-maintainers:
+    - joemarshall
diff --git a/pyarrow/patches/autoload_timezones.patch b/pyarrow/patches/autoload_timezones.patch
new file mode 100644
index 0000000..9e239dc
--- /dev/null
+++ b/pyarrow/patches/autoload_timezones.patch
@@ -0,0 +1,13 @@
+diff --git a/python/pyarrow/__init__.py.orig b/python/pyarrow/__init__.py
+index e52e0d24..9982f41c 100644
+--- a/python/pyarrow/__init__.py
++++ b/python/pyarrow/__init__.py
+@@ -28,7 +28,7 @@ streaming messaging and interprocess communication.
+ 
+ For more information see the official page at https://arrow.apache.org
+ """
+-
++import unix_timezones
+ import gc as _gc
+ import importlib as _importlib
+ import os as _os
diff --git a/scripts/get_pyodide_versions.sh b/scripts/get_pyodide_versions.sh
new file mode 100644
index 0000000..47da0f9
--- /dev/null
+++ b/scripts/get_pyodide_versions.sh
@@ -0,0 +1,6 @@
+#!/bin/bash
+set -e
+
+PACKAGE_JSON_URL="https://pypi.org/pypi/pyodide-build/json"
+
+wget --no-check-certificate "$PACKAGE_JSON_URL" -O-    | jq  -r '.releases   | keys | .[]' |sort -V