docling-parse/.github/workflows/wheels.yml

on:
  workflow_call:
    inputs:
      publish:
        type: boolean
        description: "If true, the packages will be published."
        default: false

concurrency:
  group: wheels-${{ github.workflow }}-${{ github.event.pull_request.number || github.ref }}-${{ inputs.publish }}
  cancel-in-progress: ${{ !inputs.publish }}

env:
  CMAKE_BUILD_TYPE: Release

jobs:
  build_sdist:
    name: Build sdist artifacts
    runs-on: ubuntu-24.04

    steps:
      - name: Checkout
        uses: actions/checkout@v4
      - name: Install uv and set the python version
        uses: astral-sh/setup-uv@v5
        with:
          python-version: ${{ matrix.python-version }}
      - name: Build sdist
        run: |
          uv build
          ls ./dist
      - name: Upload artifacts
        uses: actions/upload-artifact@v4
        with:
          name: python-package-distributions-sdist
          path: dist/

  build_wheels:
    name: Build wheel for py${{ matrix.python-version }} ${{ matrix.os.platform_id }}
    runs-on: ${{ matrix.os.name }}

    strategy:
      # Ensure that a wheel builder finishes even if another fails
      # list of github vm: https://docs.github.com/en/actions/using-github-hosted-runners/about-github-hosted-runners/about-github-hosted-runners#standard-github-hosted-runners-for-public-repositories
      fail-fast: false
      matrix:
        python-version: ["3.10", "3.11", "3.12", "3.13", "3.14"]

        os:
          - name: "ubuntu-24.04"
            platform: "linux"
            platform_id: "manylinux_x86_64"

          - name: "ubuntu-24.04-arm"
            platform: "linux"
            platform_id: "manylinux_aarch64"

          - name: "macos-14"
            platform: "macos"
            min_macos_version: "14"
            platform_id: "macosx_arm64"

          - name: "windows-2025"
            platform: "windows"
            platform_id: "win_amd64"      

    steps:
      - name: Checkout
        uses: actions/checkout@v4

      - name: Install uv and set the python version
        uses: astral-sh/setup-uv@v5
        id: py
        with:
          python-version: ${{ matrix.python-version }}

      - name: Set up custom PATH and set py version to cpXYZ [windows]
        if: ${{matrix.os.platform_id == 'win_amd64'}}
        shell: pwsh
        run: |
          echo "C:\msys64\usr\bin" | Out-File -FilePath $env:GITHUB_PATH -Encoding utf8 -Append
          echo "C:\msys64\mingw64\bin" | Out-File -FilePath $env:GITHUB_PATH -Encoding utf8 -Append
          echo "C:\Users\runneradmin\.local\bin" | Out-File -FilePath $env:GITHUB_PATH -Encoding utf8 -Append
          echo "C:\nasm" | Out-File -FilePath $env:GITHUB_PATH -Encoding utf8 -Append
          $version = "${{ matrix.python-version }}"
          $cp_version = "cp$($version -replace '\.', '')"
          Add-Content -Path $env:GITHUB_ENV -Value "python_cp_version=$cp_version"

      - name: Convert python version to cpXYZ
        if: ${{matrix.os.platform_id != 'win_amd64'}}
        run: |
          version=${{ matrix.python-version }}
          cp_version="cp${version//.}"
          echo "python_cp_version=$cp_version" >> $GITHUB_ENV

      - name: Build wheels [macos-x86_64]
        if: ${{matrix.os.platform_id == 'macosx_x86_64'}}
        env:
          CIBW_BUILD: ${{ env.python_cp_version }}-${{ matrix.os.platform_id }}
          CIBW_ARCHS: x86_64
          CIBW_ARCHS_MACOS: x86_64
          CIBW_PLATFORM: macos
          CIBW_SKIP: "pp* *-musllinux_* *_i686* *_s390* *arm64* *pypy*"
          CIBW_PROJECT_REQUIRES_PYTHON: "~=${{ matrix.python-version }}.0"
          CIBW_BUILD_VERBOSITY: 3
          CMAKE_OSX_ARCHITECTURES: x86_64
          CIBW_REPAIR_WHEEL_COMMAND_MACOS: ""  # do not run delocate-wheel before the re-tag
          CIBW_ENVIRONMENT: "MACOSX_DEPLOYMENT_TARGET=${{ matrix.os.min_macos_version }}.0"
          ARCHFLAGS: -arch x86_64
          BUILD_THREADS: "1"
          PYTORCH_MPS_HIGH_WATERMARK_RATIO: "0.0"
        run: |
          PY_CACHE_TAG=$(uv run python -c 'import sys;print(sys.implementation.cache_tag)')
          echo "Building wheel ${CIBW_BUILD} ${{ env.CIBW_BUILD }}"
          echo "Building cp: ${{ env.python_cp_version }}"
          echo "Building cache_tag: ${PY_CACHE_TAG}"
          echo "Building platform_id: ${{ matrix.os.platform_id }}"
          echo "Building with BUILD_THREADS: ${BUILD_THREADS:-not set}"
          uv run python --version
          uv run python --version | grep ${{ matrix.python-version }}
          cat ./pyproject.toml
          uv pip install --group build
          rm -rf ./build || true
          python -m cibuildwheel --output-dir ./wheelhouse
          echo "step 1"
          ls -l wheelhouse
          uv run wheel tags --remove --platform-tag macosx_${{ matrix.os.min_macos_version }}_0_x86_64 ./wheelhouse/*.whl
          rm -f ./wheelhouse/*arm64.whl
          echo "step 2"
          ls -l wheelhouse
          uv run delocate-wheel --require-archs x86_64 -v ./wheelhouse/*.whl
          echo "step 3"
          ls -l wheelhouse
          for file in ./wheelhouse/*.whl; do
            echo "Inspecting $file"
            uv run python -m zipfile --list "$file"
            echo "Checking if .so is contained in the wheel"
            uv run python -m zipfile --list "$file" | grep \\.so
            echo "Checking if the correct python version is contained in the wheel"
            uv run python -m zipfile --list "$file" | grep ${PY_CACHE_TAG}
          done
          mkdir -p ./dist
          cp wheelhouse/*.whl ./dist/

      # there is an error with the tagging of wheels for macosx-arm64
      # see note: https://cibuildwheel.readthedocs.io/en/stable/faq/
      # see here: https://gist.github.com/anderssonjohan/49f07e33fc5cb2420515a8ac76dc0c95#file-build-pendulum-wheels-yml-L39-L53
      - name: Build wheels [macos-arm64]
        if: ${{matrix.os.platform_id == 'macosx_arm64'}}
        env:
          CIBW_BUILD: ${{ env.python_cp_version }}-${{ matrix.os.platform_id }}
          CIBW_ARCHS: arm64
          CIBW_ARCHS_MACOS: arm64
          CIBW_PLATFORM: macos
          CIBW_SKIP: "pp* *-musllinux_* *_i686* *x86_64* *_s390* *pypy*"
          CIBW_PROJECT_REQUIRES_PYTHON: "~=${{ matrix.python-version }}.0"
          CIBW_BUILD_VERBOSITY: 3
          CMAKE_OSX_ARCHITECTURES: arm64
          CIBW_REPAIR_WHEEL_COMMAND_MACOS: ""  # do not run delocate-wheel before the re-tag
          CIBW_ENVIRONMENT: "MACOSX_DEPLOYMENT_TARGET=${{ matrix.os.min_macos_version }}.0"
          ARCHFLAGS: -arch arm64
          BUILD_THREADS: "4"
          PYTORCH_MPS_HIGH_WATERMARK_RATIO: "0.0"
          CUDA_VISIBLE_DEVICES: "cpu"
        run: |
          PY_CACHE_TAG=$(uv run python -c 'import sys;print(sys.implementation.cache_tag)')
          echo "Building wheel ${CIBW_BUILD} ${{ env.CIBW_BUILD }}"
          echo "Building cp: ${{ env.python_cp_version }}"
          echo "Building cache_tag: ${PY_CACHE_TAG}"
          echo "Building platform_id: ${{ matrix.os.platform_id }}"
          echo "Building with BUILD_THREADS: ${BUILD_THREADS:-not set}"
          uv run python --version
          uv run python --version | grep ${{ matrix.python-version }}
          cat ./pyproject.toml
          uv pip install --group build
          rm -rf ./build || true
          python -m cibuildwheel --output-dir ./wheelhouse
          echo "step 1"
          ls -l wheelhouse
          uv run wheel tags --remove --platform-tag macosx_${{ matrix.os.min_macos_version }}_0_arm64 ./wheelhouse/*.whl
          rm -f ./wheelhouse/*x86_64.whl
          echo "step 2"
          ls -l wheelhouse
          uv run delocate-wheel --require-archs arm64 -v ./wheelhouse/*.whl
          echo "step 3"
          ls -l wheelhouse
          for file in ./wheelhouse/*.whl; do
            echo "Inspecting $file"
            uv run python -m zipfile --list "$file"
            echo "Checking if .so is contained in the wheel"
            uv run python -m zipfile --list "$file" | grep \\.so
            echo "Checking if the correct python version is contained in the wheel"
            uv run python -m zipfile --list "$file" | grep ${PY_CACHE_TAG}
          done
          mkdir -p ./dist
          cp wheelhouse/*.whl ./dist/

      - name: Set up QEMU [linux]
        if: startsWith(matrix.os.name, 'ubuntu')
        uses: docker/setup-qemu-action@v3
        with:
            platforms: all
        
      - name: Build wheels [linux]
        if: startsWith(matrix.os.name, 'ubuntu')
        env:
          CIBW_BUILD: ${{ env.python_cp_version }}-${{ matrix.os.platform_id }}
          CIBW_ARCHS: auto x86_64 aarch64
          CIBW_PLATFORM: linux
          CIBW_SKIP: "pp* *musllinux_* *_i686* *_s390* *pypy*"
          CIBW_PROJECT_REQUIRES_PYTHON: "~=${{ matrix.python-version }}.0"
          CIBW_BUILD_VERBOSITY: 3
          # Multi-threading: pass BUILD_THREADS into the cibuildwheel Docker container.
          # Comment out the next line to disable parallel builds.
          CIBW_ENVIRONMENT_PASS_LINUX: "BUILD_THREADS"
          BUILD_THREADS: "4"
        run: |
          PY_CACHE_TAG=$(uv run python -c 'import sys;print(sys.implementation.cache_tag)')
          echo "Building cp: ${{ env.python_cp_version }}"
          echo "Building cache_tag: ${PY_CACHE_TAG}"
          echo "Building platform_id: ${{ matrix.os.platform_id }}"
          echo "Building with BUILD_THREADS: ${BUILD_THREADS:-not set}"
          uv run python --version
          uv run python --version | grep ${{ matrix.python-version }}
          cat ./pyproject.toml
          uv pip install --group build
          rm -rf ./build || true
          python -m cibuildwheel --output-dir ./wheelhouse
          ls -l ./wheelhouse
          for file in ./wheelhouse/*.whl; do
            echo "Inspecting $file"
            uv run python -m zipfile --list "$file"
            echo "Checking if .so is contained in the wheel"
            uv run python -m zipfile --list "$file" | grep \\.so
            echo "Checking if the correct python version is contained in the wheel"
            uv run python -m zipfile --list "$file" | grep ${PY_CACHE_TAG}
          done
          mkdir -p ./dist
          cp wheelhouse/*.whl ./dist/

      - name: Set up MSYS2 [windows]
        if: ${{matrix.os.platform_id == 'win_amd64'}}
        uses: msys2/setup-msys2@v2
        with:
          update: true
          install: >
            mingw-w64-x86_64-toolchain 
            mingw-w64-x86_64-gcc-libs
            
      - name: Set up external-libs [windows]
        if: ${{matrix.os.platform_id == 'win_amd64'}}
        shell: pwsh
        run: |
          Copy-Item -Path "C:/mingw64/bin/libgcc_s_seh-1.dll" -Destination ".\docling_parse"
          Copy-Item -Path "C:/mingw64/bin/libstdc++-6.dll" -Destination ".\docling_parse"
          Copy-Item -Path "C:/mingw64/bin/libwinpthread-1.dll" -Destination ".\docling_parse"
          New-Item -Path 'C:\windows-libs' -ItemType Directory -Force
          Invoke-WebRequest -Uri 'https://github.com/qpdf/external-libs/releases/download/release-2024-06-07/qpdf-external-libs-bin.zip' -OutFile 'C:\windows-libs\qpdf-external-libs-bin.zip'
          Expand-Archive -Path 'C:\windows-libs\qpdf-external-libs-bin.zip' -DestinationPath 'C:\windows-libs' -Force
          Get-ChildItem -Path 'C:\windows-libs' -Recurse
          
      - name: Set up NASM assembler [windows]
        if: ${{matrix.os.platform_id == 'win_amd64'}}
        shell: pwsh
        run: |
          New-Item -Path 'C:\nasm' -ItemType Directory -Force
          Invoke-WebRequest -Uri 'https://www.nasm.us/pub/nasm/releasebuilds/2.16.03/win64/nasm-2.16.03-win64.zip' -OutFile 'C:\nasm-2.16.03-win64.zip'
          Expand-Archive -Path 'C:\nasm-2.16.03-win64.zip' -DestinationPath 'C:\nasm-temp' -Force
          Copy-Item -Path 'C:\nasm-temp\nasm-2.16.03\nasm.exe' -Destination 'C:\nasm\nasm.exe'
          Remove-Item -Path 'C:\nasm-2.16.03-win64.zip' -Force
          Remove-Item -Path 'C:\nasm-temp' -Recurse -Force
          nasm -v

      - name: Build wheels
        if: ${{matrix.os.platform_id == 'win_amd64'}}
        env:
          CIBW_PLATFORM: windows
          CIBW_BUILD_VERBOSITY: 3
          CIBW_ARCHS: AMD64
          CIBW_SKIP: "pp*"
          CIBW_PROJECT_REQUIRES_PYTHON: "~=${{ matrix.python-version }}.0"
          PKG_CONFIG_PATH: "C:/msys64/usr/lib/pkgconfig"
          PKG_CONFIG_EXECUTABLE: "C:/mingw64/bin/pkg-config.exe"
          CMAKE_PREFIX_PATH: "C:/msys64/mingw64;C:/mingw64;C:/windows-libs/external-libs"
          CMAKE_LIBRARY_PATH: "C:/msys64/mingw64/lib;C:/mingw64/lib;C:/windows-libs/external-libs/lib-mingw64"
          CMAKE_INCLUDE_PATH: "C:/msys64/mingw64/include;C:/mingw64/include;C:/windows-libs/external-libs/include"
          CMAKE_GENERATOR: "MSYS Makefiles"
          BUILD_THREADS: 4
          ASM_NASM: "C:/nasm/nasm.exe"
        shell: pwsh
        run: |
          if (Test-Path "C:\Strawberry\") {
              Remove-Item -Recurse -Force "C:\Strawberry\"
          }
          $env:CMAKE_ARGS = "-DZLIB_LIBRARY=C:/windows-libs/external-libs/lib-mingw64/libz.a -DZLIB_INCLUDE_DIR=C:/windows-libs/external-libs/include -DJPEG_LIBRARY=C:/windows-libs/external-libs/lib-mingw64/libjpeg.a -DJPEG_INCLUDE_DIR=C:/windows-libs/external-libs/include"
          uv pip install --group build
          if (Test-Path ".\build") {
              Remove-Item -Recurse -Force ".\build"
          }
          uv run python -m cibuildwheel --output-dir .\wheelhouse
          Get-ChildItem -Path .\wheelhouse -Filter *.whl | Format-List
          Get-ChildItem -Path .\wheelhouse -Filter *.whl | ForEach-Object {
              $file = $_.FullName
              Write-Output "Inspecting $file"
              uv run python -m zipfile --list "$file"
          }
          if (-not (Test-Path -Path .\dist)) {
              New-Item -Path .\dist -ItemType Directory
          }
          Copy-Item -Path .\wheelhouse\*.whl -Destination .\dist\

      - name: Upload artifacts
        uses: actions/upload-artifact@v4
        with:
          name: python-package-distributions-py${{ matrix.python-version }}-${{ matrix.os.platform_id }}
          path: dist/

  publish-packages:
    name: >-
      Publish Python 🐍 distribution 📦 to PyPI
    if: inputs.publish && startsWith(github.ref, 'refs/tags/')  # only publish to PyPI on tag pushes
    needs:
      - build_sdist
      - build_wheels
    runs-on: ubuntu-24.04
    environment:
      name: pypi
      url: https://pypi.org/p/docling-parse
    permissions:
      id-token: write  # IMPORTANT: mandatory for trusted publishing
      contents: write  # IMPORTANT: mandatory for adding artifacts to GitHub Releases
    steps:
      - name: Download all the dists
        uses: actions/download-artifact@v4
        with:
          merge-multiple: true
          path: dist/
      - name: List dist/
        run: |
          ls ./dist
      - name: Remove non-manylinux Linux wheels if found
        run: |
          rm -f ./dist/*-linux_x86_64.whl
      - name: List dist/
        run: |
          ls ./dist
      - name: Sign the dists with Sigstore
        uses: sigstore/gh-action-sigstore-python@v3.0.0
        with:
          release-signing-artifacts: false
          inputs: >-
            ./dist/*.tar.gz
            ./dist/*.whl
      - name: List dist/
        run: |
          ls ./dist
      - name: Upload artifact signatures to GitHub Release
        env:
          GITHUB_TOKEN: ${{ github.token }}
        # Upload to GitHub Release using the `gh` CLI.
        # `dist/` contains the built packages, and the
        # sigstore-produced signatures and certificates.
        run: >-
          gh release upload
          '${{ github.ref_name }}' dist/**
          --repo '${{ github.repository }}'
      # PyPI does not accept .sigstore artifacts and
      # gh-action-pypi-publish has no option to ignore them.
      - name: Remove sigstore signatures before uploading to PyPI
        run: rm ./dist/*.sigstore.json
      - name: Publish distribution 📦 to PyPI
        uses: pypa/gh-action-pypi-publish@release/v1
        with:
          # currently not working with reusable workflows
          attestations: false