fix: python version in wheels

.github/actions/setup-poetry/action.yml

@@ -10,13 +10,16 @@ runs:
     - name: Install poetry
       run: pipx install poetry==1.8.3
       shell: bash
-    - uses: actions/setup-python@v4
+    - uses: actions/setup-python@v5
+      id: py
       with:
         python-version: ${{ inputs.python-version }}
         update-environment: false
         cache: 'poetry'
     - name: Setup poetry env with correct python
-      run: poetry env use python3
+      run: |
+        poetry env use ${{ steps.py.outputs.python-path }}
+        poetry run python --version
       shell: bash
     - name: Install only dependencies and not the package itself
       run: poetry install --all-extras --no-root

.github/workflows/wheels.yml

@@ -14,7 +14,7 @@ jobs:
       # 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: ["3.9", "3.10", "3.11", "3.12"]
+        python-version: ["3.10", "3.11", "3.12"]
 
         os:
           - name: "ubuntu-latest"
@@ -51,11 +51,21 @@ jobs:
 
       - name: Setup Python
         uses: actions/setup-python@v5
+        id: py
         with:
           python-version: ${{ matrix.python-version }}
+          update-environment: false
 
       - name: Install Poetry
-        run: python -m pip install poetry==1.8.3
+        run: |
+          which python
+          python --version
+          which python3
+          python3 --version
+          echo "pythonpath: ${{ steps.py.outputs.python-path }}"
+          ${{ steps.py.outputs.python-path }} --version
+          pipx install poetry==1.8.3
+          poetry env use ${{ steps.py.outputs.python-path }}
 
       - name: Set up custom PATH and set py version to cpXYZ [windows]
         if: ${{matrix.os.platform_id == 'win_amd64'}}
@@ -97,9 +107,15 @@ jobs:
           CIBW_ENVIRONMENT: "MACOSX_DEPLOYMENT_TARGET=${{ matrix.os.macos_version }}.0"
           ARCHFLAGS: -arch x86_64
           BUILD_THREADS: "4"
+          PYTORCH_MPS_HIGH_WATERMARK_RATIO: "0.0"
         run: |
-          echo "Building wheel ${CIBW_BUILD}"
+          PY_CACHE_TAG=$(poetry 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 }}"
           poetry run python --version
+          poetry run python --version | grep ${{ matrix.python-version }} 
           poetry install --no-root --only=build
           cat ./pyproject.toml
           poetry run python -m cibuildwheel --output-dir wheelhouse
@@ -115,6 +131,10 @@ jobs:
           for file in ./wheelhouse/*.whl; do
             echo "Inspecting $file"
             poetry run python -m zipfile --list "$file"
+            echo "Checking if .so is contained in the wheel"
+            poetry run python -m zipfile --list "$file" | grep \\.so
+            echo "Checking if the correct python version is contained in the wheel"
+            poetry run python -m zipfile --list "$file" | grep ${PY_CACHE_TAG}
           done
           mkdir -p ./dist
           cp wheelhouse/*.whl ./dist/
@@ -137,9 +157,16 @@ jobs:
           CIBW_ENVIRONMENT: "MACOSX_DEPLOYMENT_TARGET=${{ matrix.os.macos_version }}.0"
           ARCHFLAGS: -arch arm64
           BUILD_THREADS: "4"
+          PYTORCH_MPS_HIGH_WATERMARK_RATIO: "0.0"
+          CUDA_VISIBLE_DEVICES: "cpu"
         run: |
-          echo "Building wheel ${CIBW_BUILD}"
+          PY_CACHE_TAG=$(poetry 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 }}"
           poetry run python --version
+          poetry run python --version | grep ${{ matrix.python-version }} 
           poetry install --no-root --only=build
           cat ./pyproject.toml
           poetry run python -m cibuildwheel --output-dir wheelhouse
@@ -155,6 +182,10 @@ jobs:
           for file in ./wheelhouse/*.whl; do
             echo "Inspecting $file"
             poetry run python -m zipfile --list "$file"
+            echo "Checking if .so is contained in the wheel"
+            poetry run python -m zipfile --list "$file" | grep \\.so
+            echo "Checking if the correct python version is contained in the wheel"
+            poetry run python -m zipfile --list "$file" | grep ${PY_CACHE_TAG}
           done
           mkdir -p ./dist
           cp wheelhouse/*.whl ./dist/
@@ -183,15 +214,23 @@ jobs:
           CIBW_BUILD_VERBOSITY: 3
           BUILD_THREADS: "8"
         run: |
-          echo "Building wheel ${CIBW_BUILD}"
+          PY_CACHE_TAG=$(poetry 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 }}"
           poetry run python --version
+          poetry run python --version | grep ${{ matrix.python-version }} 
           poetry install --no-root --only=build
           cat ./pyproject.toml
           poetry run python -m cibuildwheel --output-dir ./wheelhouse
           ls -l ./wheelhouse
           for file in ./wheelhouse/*.whl; do
             echo "Inspecting $file"
             poetry run python -m zipfile --list "$file"
+            echo "Checking if .so is contained in the wheel"
+            poetry run python -m zipfile --list "$file" | grep \\.so
+            echo "Checking if the correct python version is contained in the wheel"
+            poetry run python -m zipfile --list "$file" | grep ${PY_CACHE_TAG}
           done
           mkdir -p ./dist
           cp wheelhouse/*.whl ./dist/
@@ -256,9 +295,9 @@ jobs:
 
       - name: publish wheels (dry run)
         run: |
-          poetry publish --dry-run --no-interaction -vvv
+          poetry publish --skip-existing --dry-run --no-interaction -vvv
 
       - name: publish wheels (on publishing) [for releases only]
         if: ${{ startsWith(github.ref, 'refs/tags/') }}
         run: |
-          poetry publish --no-interaction -vvv
+          poetry publish --skip-existing --no-interaction -vvv

CMakeLists.txt

@@ -1,4 +1,4 @@
-cmake_minimum_required(VERSION 3.12)
+cmake_minimum_required(VERSION 3.12..3.26)
 
 project(docling_parse VERSION 1.0.0 LANGUAGES CXX C)
 #set(CMAKE_VERBOSE_MAKEFILE off)
@@ -147,7 +147,7 @@ target_link_libraries(parse ${LIB_LINK})
 # ***************************
 
 # https://pybind11.readthedocs.io/en/stable/compiling.html
-find_package (Python COMPONENTS Interpreter Development)
+# https://pybind11.readthedocs.io/en/stable/compiling.html#configuration-variables
 find_package(pybind11 CONFIG REQUIRED)
 
 pybind11_add_module(docling_parse "${TOPLEVEL_PREFIX_PATH}/app/pybind_parse.cpp")

build.py

@@ -41,11 +41,11 @@ def run(cmd: List[str], cwd: str="./"):
 def build_local(num_threads: int):
 
     print("python prefix: ", sys.exec_prefix)
+    print("python executable: ", sys.executable)
     config_cmd = [
         "cmake",
         "-B", f"{BUILD_DIR}",
-        f"-DPython_ROOT_DIR={sys.exec_prefix}",
-        f"-DPython3_ROOT_DIR={sys.exec_prefix}",
+        f"-DPYTHON_EXECUTABLE={sys.executable}",
     ]
     config_cmd.extend(get_pybind11_cmake_args())
     success = run(config_cmd, cwd=ROOT_DIR)