fix merge conflicts with main

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

@@ -10,10 +10,17 @@ 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 ${{ 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
       shell: bash

.github/scripts/build_rhel.sh

@@ -0,0 +1,34 @@
+#!/bin/bash
+
+set -e  # trigger failure on error - do not remove!
+set -x  # display command on output
+
+# Build the sdist
+poetry build -f sdist
+
+# Compile the wheel from sdist in centos stream
+
+docker build -f - . <<EOF
+    FROM quay.io/centos/centos:stream9
+    RUN dnf config-manager --set-enabled crb
+    # RUN dnf copr -y enable cheimes/deepsearch-glm rhel-9-x86_64
+    RUN dnf install -y https://dl.fedoraproject.org/pub/epel/epel-release-latest-9.noarch.rpm \
+       && dnf clean all
+    RUN dnf install -y --nodocs \
+            gcc gcc-c++ git make cmake pkgconfig glibc-devel \
+            python3.11 python3.11-pip python3.11-devel \
+            libjpeg-turbo-devel libpng-devel qpdf-devel json-devel utf8cpp-devel zlib-devel \
+            loguru-devel \
+        && dnf clean all
+
+    RUN mkdir /src
+    COPY ./dist/*.tar.gz /src/
+
+    RUN USE_SYSTEM_DEPS=ON pip3.11 install /src/docling_parse*.tar.gz \
+        && python3.11 -c 'from docling_parse.docling_parse import pdf_parser, pdf_parser_v2'
+
+    COPY ./tests /src/tests
+    RUN cd /src \
+        && pip3.11 install pytest \
+        && pytest
+EOF

.github/workflows/checks.yml

@@ -12,16 +12,16 @@ jobs:
         python-version: ['3.9', '3.10', '3.11', '3.12']
     steps:
       - uses: actions/checkout@v4
-      # - name: Install dependencies [linux]
-      #   run: sudo apt-get install -y libldap-common
-      #   shell: bash
       - uses: ./.github/actions/setup-poetry
         with:
           python-version: ${{ matrix.python-version }}
       - name: Run styling check
         run: poetry run pre-commit run --all-files
       - name: Install with poetry
-        run: poetry install --all-extras
+        run: |
+          poetry install --all-extras
+          ls -l
+          ls -l docling_parse
       - name: Testing
         run: |
           poetry run pytest -v tests

.github/workflows/ci.yml

@@ -21,4 +21,6 @@ jobs:
     uses: ./.github/workflows/checks.yml
   build-wheels:
     uses: ./.github/workflows/wheels.yml
+  rhel-build:
+    uses: ./.github/workflows/rhel.yml
 

.github/workflows/rhel.yml

@@ -0,0 +1,12 @@
+on:
+  workflow_call:
+
+jobs:
+  run-checks:
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+      - uses: ./.github/actions/setup-poetry
+      - name: Run build in docker
+        run: ./.github/scripts/build_rhel.sh
+        shell: bash

.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
@@ -115,8 +131,12 @@ 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 ./dist
+          mkdir -p ./dist
           cp wheelhouse/*.whl ./dist/
 
       # there is an error with the tagging of wheels for macosx-arm64
@@ -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
@@ -155,8 +182,12 @@ 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 ./dist
+          mkdir -p ./dist
           cp wheelhouse/*.whl ./dist/
 
       - name: Set up QEMU [linux]
@@ -165,6 +196,13 @@ jobs:
         with:
             platforms: all
 
+      - name: Build sdist
+        # build only on Linux to avoid too many duplicates of the sdist
+        if: matrix.os.name == 'ubuntu-latest' 
+        run: |
+          echo "Building wheel ${CIBW_BUILD}"
+          poetry build -f sdist
+
       - name: Build wheels [linux]
         if: matrix.os.name == 'ubuntu-latest'
         env:
@@ -176,17 +214,25 @@ 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 ./dist
+          mkdir -p ./dist
           cp wheelhouse/*.whl ./dist/
 
       - name: Set up MSYS2 [windows]
@@ -196,12 +242,15 @@ jobs:
           update: true
           install: >
             mingw-w64-x86_64-toolchain 
-            mingw-w64-i686-toolchain
+            mingw-w64-x86_64-gcc-libs
 
-      - name: Set up QPDF external-libs [windows]
+      - name: Set up external-libs [windows]
         if: ${{matrix.os.platform_id == 'win_amd64'}}
         shell: pwsh
         run: |
+          Copy-Item -Path "C:/mingw64/bin/libgcc_s_seh-1.dll" -Destination ".\docling_parse"
+          Copy-Item -Path "C:/mingw64/bin/libstdc++-6.dll" -Destination ".\docling_parse"
+          Copy-Item -Path "C:/mingw64/bin/libwinpthread-1.dll" -Destination ".\docling_parse"
           New-Item -Path 'C:\windows-libs' -ItemType Directory -Force
           Invoke-WebRequest -Uri 'https://github.com/qpdf/external-libs/releases/download/release-2024-06-07/qpdf-external-libs-bin.zip' -OutFile 'C:\windows-libs\qpdf-external-libs-bin.zip'
           Expand-Archive -Path 'C:\windows-libs\qpdf-external-libs-bin.zip' -DestinationPath 'C:\windows-libs' -Force
@@ -222,11 +271,11 @@ jobs:
           CIBW_BUILD_VERBOSITY: 3
           CIBW_ARCHS: AMD64
           CIBW_PROJECT_REQUIRES_PYTHON: "~=${{ matrix.python-version }}.0"
-          PKG_CONFIG_PATH: "C:/msys64/mingw64/lib/pkgconfig"
-          PKG_CONFIG_EXECUTABLE: "C:/msys64/usr/bin/pkg-config.exe"
-          CMAKE_PREFIX_PATH: "C:/msys64/mingw64;C:/windows-libs/external-libs"
-          CMAKE_LIBRARY_PATH: "C:/msys64/mingw64/lib;C:/windows-libs/external-libs/lib-mingw64"
-          CMAKE_INCLUDE_PATH: "C:/msys64/mingw64/include;C:/windows-libs/external-libs/include"
+          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: 1
           ASM_NASM: "C:/nasm/nasm.exe"
@@ -249,9 +298,9 @@ 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

.gitignore

@@ -287,7 +287,6 @@ pyrightconfig.json
 [Ll]ib
 [Ll]ib64
 [Ll]ocal
-[Ss]cripts
 pyvenv.cfg
 pip-selfcheck.json
 

CHANGELOG.md

@@ -1,3 +1,25 @@
+## [v1.4.1](https://github.com/DS4SD/docling-parse/releases/tag/v1.4.1) - 2024-10-02
+
+### Fix
+
+* Windows build properly linking to system libraries ([#36](https://github.com/DS4SD/docling-parse/issues/36)) ([`e26ed05`](https://github.com/DS4SD/docling-parse/commit/e26ed056c22400552918c3a97dfb13614c9a03f5))
+
+## [v1.4.0](https://github.com/DS4SD/docling-parse/releases/tag/v1.4.0) - 2024-10-02
+
+### Feature
+
+* Build using system deps ([#33](https://github.com/DS4SD/docling-parse/issues/33)) ([`e1c8e49`](https://github.com/DS4SD/docling-parse/commit/e1c8e4980faab35bfdf6d1a78d8749745c560889))
+
+### Fix
+
+* Python version in wheels ([#31](https://github.com/DS4SD/docling-parse/issues/31)) ([`8d903ba`](https://github.com/DS4SD/docling-parse/commit/8d903baf61a7706066374c23265e115a9513c3ba))
+
+## [v1.3.1](https://github.com/DS4SD/docling-parse/releases/tag/v1.3.1) - 2024-09-30
+
+### Fix
+
+* Sdist and wheels content ([#28](https://github.com/DS4SD/docling-parse/issues/28)) ([`f3febc5`](https://github.com/DS4SD/docling-parse/commit/f3febc53a2a6565b16847113633f92d1a2dab48a))
+
 ## [v1.3.0](https://github.com/DS4SD/docling-parse/releases/tag/v1.3.0) - 2024-09-20
 
 ### Feature