⬆️ upgrade dependencies and add more tools

Author: sebastianMindee

.github/dependabot.yml

@@ -0,0 +1,70 @@
+version: 2
+
+updates:
+  - package-ecosystem: "pip"
+    directory: "/"
+    schedule:
+      interval: "weekly"
+      day: "monday"
+      time: "09:00"
+      timezone: "Europe/Paris"
+    open-pull-requests-limit: 8
+    commit-message:
+      prefix: "chore(deps)"
+    groups:
+      runtime-dependencies:
+        patterns:
+          - "pytz"
+          - "requests"
+      optional-dependencies:
+        patterns:
+          - "pypdfium2"
+          - "Pillow"
+      test-dependencies:
+        patterns:
+          - "pytest"
+          - "pytest-cov"
+          - "toml"
+      lint-dependencies:
+        patterns:
+          - "pylint"
+          - "pre-commit"
+          - "types-*"
+      docs-dependencies:
+        patterns:
+          - "sphinx"
+          - "sphinx-*"
+      build-dependencies:
+        patterns:
+          - "build"
+          - "twine"
+
+  - package-ecosystem: "github-actions"
+    directory: "/"
+    schedule:
+      interval: "weekly"
+      day: "monday"
+      time: "09:30"
+      timezone: "Europe/Paris"
+    open-pull-requests-limit: 3
+    commit-message:
+      prefix: "chore(ci)"
+    groups:
+      github-actions:
+        patterns:
+          - "*"
+
+  - package-ecosystem: "pre-commit"
+    directory: "/"
+    schedule:
+      interval: "weekly"
+      day: "monday"
+      time: "10:00"
+      timezone: "Europe/Paris"
+    open-pull-requests-limit: 3
+    commit-message:
+      prefix: "chore(pre-commit)"
+    groups:
+      pre-commit-hooks:
+        patterns:
+          - "*"

.github/workflows/_publish-docs.yml

@@ -39,8 +39,14 @@ jobs:
         cd ./docs
         make html
 
+    - name: Sphinx make linkcheck
+      run: |
+        cd ./docs
+        make linkcheck
+
     - name: Deploy
       uses: peaceiris/actions-gh-pages@v4
+      if: success()
       with:
         github_token: ${{ secrets.GITHUB_TOKEN }}
         publish_dir: ./docs/_build/html

.pre-commit-config.yaml

@@ -6,13 +6,13 @@ repos:
         args: [ --fix, --exit-non-zero-on-fix]
       - id: ruff-format
 
-  - repo: https://github.com/gitleaks/gitleaks
-    rev: v8.30.0
+  - repo: https://github.com/betterleaks/betterleaks
+    rev: v1.3.1
     hooks:
-      - id: gitleaks
+      - id: betterleaks
 
   - repo: https://github.com/PyCQA/pylint
-    rev: v3.3.9
+    rev: v4.0.5
     hooks:
       - id: pylint
         name: pylint
@@ -22,6 +22,11 @@ repos:
         args: [
           "-j2"
         ]
+  - repo: https://github.com/pypa/pip-audit
+    rev: v2.10.0
+    hooks:
+      - id: pip-audit
+        args: ["."]
 
   - repo: https://github.com/pre-commit/mirrors-mypy
     rev: v1.18.2

mindee/image/image_extractor.py

@@ -21,7 +21,6 @@ def attach_image_as_new_file(  # type: ignore
     :params input_buffer: Input buffer.
     :return: A PdfDocument handle.
     """
-    # Create a new page in the PdfDocument
     input_buffer.seek(0)
     image = Image.open(input_buffer)
     image.convert("RGB")
@@ -32,8 +31,7 @@ def attach_image_as_new_file(  # type: ignore
 
     image_pdf = pdfium.PdfImage.new(pdf)
     image_pdf.load_jpeg(image_buffer)
-    width, height = image_pdf.get_size()
-
+    width, height = image.width, image.height
     matrix = pdfium.PdfMatrix().scale(width, height)
     image_pdf.set_matrix(matrix)
 

mindee/pdf/pdf_compressor.py

@@ -213,7 +213,8 @@ def _collect_images_as_pdf(image_list: List[bytes]) -> pdfium.PdfDocument:  # ty
         pdf_image = pdfium.PdfImage.new(out_pdf)
         pdf_image.load_jpeg(io.BytesIO(image_bytes))
 
-        width, height = pdf_image.get_size()
+        metadata = pdf_image.get_metadata()
+        width, height = metadata.width, metadata.height
         page = out_pdf.new_page(width, height)
         page.insert_obj(pdf_image)
         page.gen_content()

pyproject.toml

@@ -28,10 +28,10 @@ classifiers = [
 requires-python = ">=3.10"
 
 dependencies = [
-  "pypdfium2>=4.0,<4.30.1",
-  "Pillow>=9.5.0",
-  "pytz>=2023.3",
-  "requests>=2.31.0",
+  "pypdfium2>=4.0,<6.0",
+  "Pillow>=12.2.0",
+  "pytz>=2026.2",
+  "requests>=2.34.2",
 ]
 
 [project.urls]
@@ -43,20 +43,21 @@ Changelog = "https://github.com/mindee/mindee-api-python/blob/main/CHANGELOG.md"
 
 [project.optional-dependencies]
 lint = [
-  "pylint==3.3.9",
-  "pre-commit~=3.6.0",
-  "types-pytz>=2024.2",
-  "types-requests>=2.31",
+  "pylint==4.0.5",
+  "pre-commit~=4.6.0",
+  "types-pytz>=2026.2.0.20260518",
+  "types-requests>=2.33.0.20260518",
+  "pip-audit>=2.10.0",
 ]
 test = [
   "toml~=0.10.2",
-  "pytest~=8.0.0",
-  "pytest-cov~=5.0",
+  "pytest~=9.0.3",
+  "pytest-cov~=7.1.0",
 ]
 docs = [
-  "sphinx~=7.3",
-  "sphinx_rtd_theme~=2.0",
-  "sphinx-autodoc-typehints~=2.2",
+  "sphinx~=9.1.0",
+  "sphinx_rtd_theme~=3.1.0",
+  "sphinx-autodoc-typehints~=3.10.4",
 ]
 build = [
   "build",

tests/v2/file_operations/test_crop_operation_integration.py

@@ -51,9 +51,10 @@ def test_image_should_extract_crops():
     )
     check_findoc_return(invoice_0)
     extracted_images.save_all_to_disk(OUTPUT_DIR)
-    # note: flaky
-    assert os.path.getsize(OUTPUT_DIR / "crop_001.jpg") in (187601, 199685)
-    assert os.path.getsize(OUTPUT_DIR / "crop_002.jpg") in (197978, 199433)
+    crop1size = os.path.getsize(OUTPUT_DIR / "crop_001.jpg")
+    crop2size = os.path.getsize(OUTPUT_DIR / "crop_002.jpg")
+    assert 187601 <= crop1size <= 199685
+    assert crop2size == 197978
 
 
 @pytest.fixture(scope="module", autouse=True)