🧪 ✨ use in-house packages for PDF & image handling

.github/dependabot.yml

@@ -17,7 +17,7 @@ updates:
           - "requests"
       optional-dependencies:
         patterns:
-          - "pypdfium2"
+          - "bernard_ledit"
           - "Pillow"
       test-dependencies:
         patterns:

.github/workflows/_build-bernard-ledit.yml

@@ -0,0 +1,89 @@
+name: Build bernard-ledit Wheel
+
+on:
+  workflow_call:
+  workflow_dispatch:
+
+permissions:
+  contents: read
+
+jobs:
+  build-wheel:
+    name: Build bernard-ledit wheel
+    timeout-minutes: 30
+    strategy:
+      fail-fast: false
+      matrix:
+        os:
+          - "ubuntu-22.04"
+          - "windows-2022"
+          - "macos-latest"
+    runs-on: ${{ matrix.os }}
+    steps:
+      - uses: actions/checkout@v6
+        with:
+          submodules: recursive
+
+      - name: Set up Python 3.10
+        uses: actions/setup-python@v6
+        with:
+          python-version: "3.10"
+
+      - name: Resolve bernard-ledit submodule commit
+        id: bernard
+        shell: bash
+        run: echo "sha=$(git -C bernard-ledit rev-parse HEAD)" >> "$GITHUB_OUTPUT"
+
+      # Skip the whole Rust/maturin build when the submodule pointer is unchanged.
+      - name: Restore cached bernard-ledit build
+        id: cache
+        uses: actions/cache@v5
+        with:
+          path: ./.github/artifacts/bernard-ledit
+          key: bernard-ledit-${{ runner.os }}-${{ steps.bernard.outputs.sha }}
+
+      - name: Cache downloaded PDFium binary
+        if: steps.cache.outputs.cache-hit != 'true'
+        uses: actions/cache@v5
+        with:
+          path: bernard-ledit/core/.pdfium_cache
+          key: pdfium-${{ runner.os }}-${{ hashFiles('bernard-ledit/core/build.rs') }}
+          restore-keys: |
+            pdfium-${{ runner.os }}-
+
+      - name: Cache Rust build
+        if: steps.cache.outputs.cache-hit != 'true'
+        uses: Swatinem/rust-cache@v2
+        with:
+          workspaces: bernard-ledit
+          key: ${{ steps.bernard.outputs.sha }}
+
+      - name: Build bernard-ledit wheel
+        if: steps.cache.outputs.cache-hit != 'true'
+        shell: bash
+        run: |
+          python -m pip install --upgrade pip
+          pip wheel --no-deps --wheel-dir ./.github/artifacts/bernard-ledit bernard-ledit/bindings/python/
+
+      # The wheel does not embed PDFium; ship the library so consumers can set PDFIUM_PATH.
+      - name: Collect bundled PDFium library
+        if: steps.cache.outputs.cache-hit != 'true'
+        shell: bash
+        run: |
+          lib=$(find bernard-ledit/core/.pdfium_cache \( -name 'libpdfium.*' -o -name 'pdfium.dll' \) -print | head -n 1)
+          if [ -z "$lib" ]; then
+            echo "::error::PDFium library not found under bernard-ledit/core/.pdfium_cache after build"
+            exit 1
+          fi
+          cp "$lib" ./.github/artifacts/bernard-ledit/
+          echo "Bundled PDFium library: $(basename "$lib")"
+
+      - name: Upload bernard-ledit build
+        uses: actions/upload-artifact@v4
+        with:
+          name: bernard-ledit-wheel-${{ runner.os }}
+          path: |
+            ./.github/artifacts/bernard-ledit/*.whl
+            ./.github/artifacts/bernard-ledit/libpdfium.*
+            ./.github/artifacts/bernard-ledit/pdfium.dll
+          if-no-files-found: error

.github/workflows/_smoke-test.yml

@@ -42,9 +42,25 @@ jobs:
           restore-keys: |
             ${{ runner.os }}-samples-
 
+      - name: Download prebuilt bernard-ledit wheel
+        uses: actions/download-artifact@v5
+        continue-on-error: true
+        with:
+          name: bernard-ledit-wheel-${{ runner.os }}
+          path: ./.github/artifacts/bernard-ledit
+
+      - name: Locate bundled PDFium
+        shell: bash
+        run: |
+          lib=$(find .github/artifacts/bernard-ledit \( -name 'libpdfium.*' -o -name 'pdfium.dll' \) -print 2>/dev/null | head -n 1)
+          if [ -n "$lib" ]; then
+            echo "PDFIUM_PATH=$(python -c 'import os,sys;print(os.path.abspath(sys.argv[1]))' "$lib")" >> "$GITHUB_ENV"
+          fi
+
       - name: Install dependencies
         run: |
           python -m pip install pip
+          python -c "from pathlib import Path; import subprocess, sys; wheels = sorted(Path('.github/artifacts/bernard-ledit').glob('*.whl')); subprocess.check_call([sys.executable, '-m', 'pip', 'install', str(wheels[0]) if wheels else 'bernard-ledit/bindings/python/'])"
           pip install -e '.'
 
       - name: Tests v2 sample code

.github/workflows/_static-analysis.yml

@@ -15,6 +15,8 @@ jobs:
         python-version: ["3.12"]
     steps:
     - uses: actions/checkout@v6
+      with:
+        submodules: recursive
 
     - name: Set up Python ${{ matrix.python-version }}
       uses: actions/setup-python@v6
@@ -29,10 +31,26 @@ jobs:
         restore-keys: |
           ${{ runner.os }}-lint-
 
+    - name: Download prebuilt bernard-ledit wheel
+      uses: actions/download-artifact@v5
+      continue-on-error: true
+      with:
+        name: bernard-ledit-wheel-${{ runner.os }}
+        path: ./.github/artifacts/bernard-ledit
+
+    - name: Locate bundled PDFium
+      shell: bash
+      run: |
+        lib=$(find .github/artifacts/bernard-ledit \( -name 'libpdfium.*' -o -name 'pdfium.dll' \) -print 2>/dev/null | head -n 1)
+        if [ -n "$lib" ]; then
+          echo "PDFIUM_PATH=$(python -c 'import os,sys;print(os.path.abspath(sys.argv[1]))' "$lib")" >> "$GITHUB_ENV"
+        fi
+
     - name: Install dependencies
       run: |
         python -m pip install pip
         pip install pylic~=3.6.1
+        python -c "from pathlib import Path; import subprocess, sys; wheels = sorted(Path('.github/artifacts/bernard-ledit').glob('*.whl')); subprocess.check_call([sys.executable, '-m', 'pip', 'install', str(wheels[0]) if wheels else 'bernard-ledit/bindings/python/'])"
         pip install -e '.'
 
     - name: License check
@@ -43,6 +61,10 @@ jobs:
       run: |
         pip install -e '.[lint]'
 
+    - name: Audit dependencies
+      run: |
+        pip-audit --local --skip-editable
+
     - name: Cache pre-commit
       uses: actions/cache@v5
       with:

.github/workflows/_test-cli.yml

@@ -48,9 +48,25 @@ jobs:
         restore-keys: |
           ${{ runner.os }}-cli-
 
+    - name: Download prebuilt bernard-ledit wheel
+      uses: actions/download-artifact@v5
+      continue-on-error: true
+      with:
+        name: bernard-ledit-wheel-${{ runner.os }}
+        path: ./.github/artifacts/bernard-ledit
+
+    - name: Locate bundled PDFium
+      shell: bash
+      run: |
+        lib=$(find .github/artifacts/bernard-ledit \( -name 'libpdfium.*' -o -name 'pdfium.dll' \) -print 2>/dev/null | head -n 1)
+        if [ -n "$lib" ]; then
+          echo "PDFIUM_PATH=$(python -c 'import os,sys;print(os.path.abspath(sys.argv[1]))' "$lib")" >> "$GITHUB_ENV"
+        fi
+
     - name: Install dependencies
       run: |
         python -m pip install pip
+        python -c "from pathlib import Path; import subprocess, sys; wheels = sorted(Path('.github/artifacts/bernard-ledit').glob('*.whl')); subprocess.check_call([sys.executable, '-m', 'pip', 'install', str(wheels[0]) if wheels else 'bernard-ledit/bindings/python/'])"
         pip install -e '.'
 
     - name: Test V1 CLI

.github/workflows/_test-integrations.yml

@@ -41,9 +41,25 @@ jobs:
         restore-keys: |
           ${{ runner.os }}-test-
 
+    - name: Download prebuilt bernard-ledit wheel
+      uses: actions/download-artifact@v5
+      continue-on-error: true
+      with:
+        name: bernard-ledit-wheel-${{ runner.os }}
+        path: ./.github/artifacts/bernard-ledit
+
+    - name: Locate bundled PDFium
+      shell: bash
+      run: |
+        lib=$(find .github/artifacts/bernard-ledit \( -name 'libpdfium.*' -o -name 'pdfium.dll' \) -print 2>/dev/null | head -n 1)
+        if [ -n "$lib" ]; then
+          echo "PDFIUM_PATH=$(python -c 'import os,sys;print(os.path.abspath(sys.argv[1]))' "$lib")" >> "$GITHUB_ENV"
+        fi
+
     - name: Install dependencies
       run: |
         python -m pip install pip
+        python -c "from pathlib import Path; import subprocess, sys; wheels = sorted(Path('.github/artifacts/bernard-ledit').glob('*.whl')); subprocess.check_call([sys.executable, '-m', 'pip', 'install', str(wheels[0]) if wheels else 'bernard-ledit/bindings/python/'])"
         pip install -e '.[test]'
     - name: Run Integration Testing
       env:
@@ -116,4 +132,4 @@ jobs:
           MINDEE_V2_SE_TESTS_SPLIT_MODEL_ID: ${{ secrets.MINDEE_V2_SE_TESTS_SPLIT_MODEL_ID }}
           MINDEE_V2_SE_TESTS_OCR_MODEL_ID: ${{ secrets.MINDEE_V2_SE_TESTS_OCR_MODEL_ID }}
       run: |
-        pytest -m "integration and not pypdfium2 and not pillow"
+        pytest -m "integration and not bernard_ledit and not pillow"

.github/workflows/_test-regressions.yml

@@ -38,9 +38,25 @@ jobs:
         restore-keys: |
           ${{ runner.os }}-test-
 
+    - name: Download prebuilt bernard-ledit wheel
+      uses: actions/download-artifact@v5
+      continue-on-error: true
+      with:
+        name: bernard-ledit-wheel-${{ runner.os }}
+        path: ./.github/artifacts/bernard-ledit
+
+    - name: Locate bundled PDFium
+      shell: bash
+      run: |
+        lib=$(find .github/artifacts/bernard-ledit \( -name 'libpdfium.*' -o -name 'pdfium.dll' \) -print 2>/dev/null | head -n 1)
+        if [ -n "$lib" ]; then
+          echo "PDFIUM_PATH=$(python -c 'import os,sys;print(os.path.abspath(sys.argv[1]))' "$lib")" >> "$GITHUB_ENV"
+        fi
+
     - name: Install dependencies
       run: |
         python -m pip install pip
+        python -c "from pathlib import Path; import subprocess, sys; wheels = sorted(Path('.github/artifacts/bernard-ledit').glob('*.whl')); subprocess.check_call([sys.executable, '-m', 'pip', 'install', str(wheels[0]) if wheels else 'bernard-ledit/bindings/python/'])"
         pip install -e '.[test]'
     - name: Run Regression Testing
       env:

.github/workflows/_test-units.yml

@@ -15,6 +15,7 @@ jobs:
         os:
           - "ubuntu-22.04"
           - "windows-2022"
+          - "macos-latest"
         python-version:
           - "3.10"
           - "3.11"
@@ -40,9 +41,25 @@ jobs:
         restore-keys: |
           ${{ runner.os }}-test-
 
+    - name: Download prebuilt bernard-ledit wheel
+      uses: actions/download-artifact@v5
+      continue-on-error: true
+      with:
+        name: bernard-ledit-wheel-${{ runner.os }}
+        path: ./.github/artifacts/bernard-ledit
+
+    - name: Locate bundled PDFium
+      shell: bash
+      run: |
+        lib=$(find .github/artifacts/bernard-ledit \( -name 'libpdfium.*' -o -name 'pdfium.dll' \) -print 2>/dev/null | head -n 1)
+        if [ -n "$lib" ]; then
+          echo "PDFIUM_PATH=$(python -c 'import os,sys;print(os.path.abspath(sys.argv[1]))' "$lib")" >> "$GITHUB_ENV"
+        fi
+
     - name: Install dependencies
       run: |
         python -m pip install pip
+        python -c "from pathlib import Path; import subprocess, sys; wheels = sorted(Path('.github/artifacts/bernard-ledit').glob('*.whl')); subprocess.check_call([sys.executable, '-m', 'pip', 'install', str(wheels[0]) if wheels else 'bernard-ledit/bindings/python/'])"
         pip install -e '.[test]'
 
     - name: Unit testing with pytest
@@ -57,6 +74,7 @@ jobs:
         os:
           - "ubuntu-22.04"
           - "windows-2022"
+          - "macos-latest"
         python-version:
           - "3.10"
           - "3.11"
@@ -92,4 +110,3 @@ jobs:
     - name: Unit testing with pytest
       run: |
         pytest
-

.github/workflows/cron.yml

@@ -10,12 +10,17 @@ permissions:
   actions: read
 
 jobs:
+  build-bernard-ledit:
+    uses: mindee/mindee-api-python/.github/workflows/_build-bernard-ledit.yml@main
   test-regressions:
     uses: mindee/mindee-api-python/.github/workflows/_test-regressions.yml@main
+    needs: build-bernard-ledit
     secrets: inherit
   test-code-samples:
     uses: mindee/mindee-api-python/.github/workflows/_smoke-test.yml@main
+    needs: build-bernard-ledit
     secrets: inherit
   test-cli:
     uses: mindee/mindee-api-python/.github/workflows/_test-cli.yml@main
+    needs: build-bernard-ledit
     secrets: inherit

.github/workflows/pull-request.yml

@@ -10,12 +10,16 @@ permissions:
 jobs:
   workflow-lint:
     uses: ./.github/workflows/_workflow_lint.yml
+  build-bernard-ledit:
+    uses: ./.github/workflows/_build-bernard-ledit.yml
   static-analysis:
     uses: ./.github/workflows/_static-analysis.yml
-    needs: workflow-lint
+    needs:
+      - workflow-lint
+      - build-bernard-ledit
   test-units:
     uses: ./.github/workflows/_test-units.yml
-    needs: static-analysis
+    needs: build-bernard-ledit
     secrets: inherit
   test-regressions:
     uses: ./.github/workflows/_test-regressions.yml

.github/workflows/push-main-branch.yml

@@ -6,15 +6,20 @@
       - main
 
 jobs:
+  build-bernard-ledit:
+    uses: mindee/mindee-api-python/.github/workflows/_build-bernard-ledit.yml@main
   static-analysis:
     uses: mindee/mindee-api-python/.github/workflows/_static-analysis.yml@main
+    needs: build-bernard-ledit
   test_units:
     uses: mindee/mindee-api-python/.github/workflows/_test-units.yml@main
-    needs: static-analysis
+    needs: build-bernard-ledit
     secrets: inherit
   tag:
     uses: mindee/client-lib-actions/.github/workflows/tag-version.yml@main
-    needs: test_units
+    needs:
+      - static-analysis
+      - test_units
   release:
     uses: mindee/client-lib-actions/.github/workflows/create-release.yml@main
     needs: tag

.gitmodules

@@ -1,3 +1,7 @@
 [submodule "tests/data"]
 	path = tests/data
 	url = git@github.com:mindee/client-lib-test-data.git
+[submodule "bernard-ledit"]
+	path = bernard-ledit
+	url = https://github.com/mindee/bernard-ledit.git
+	branch = main

.pre-commit-config.yaml

@@ -22,14 +22,6 @@ repos:
         args: [
           "-j2"
         ]
-  - repo: https://github.com/pypa/pip-audit
-    rev: v2.10.1
-    hooks:
-      - id: pip-audit
-        args: ["."]
-        files: ^(requirements.*\.txt|setup\.cfg|setup\.py|pyproject\.toml)$
-        stages: [pre-push]
-
   - repo: https://github.com/pre-commit/mirrors-mypy
     rev: v2.1.0
     hooks: