Labelformat: Add python 3.14 support

.github/workflows/run-tests.yml

@@ -14,7 +14,7 @@ jobs:
     runs-on: ubuntu-22.04
     strategy:
       matrix:
-        python: ["3.8", "3.10"]
+        python: ["3.8", "3.14"]
 
     steps:
       - name: Checkout code

README.md

@@ -44,7 +44,7 @@ and time-consuming. Labelformat aims to solve this pain.
 
 - Support for common dataset label formats (more coming soon)
 - Support for common tool formats (more coming soon)
-- Minimal dependencies, targets python 3.8 or higher
+- Minimal dependencies, supports Python 3.8 through 3.14
 - Memory concious - datasets are processed file-by-file instead of loading everything
   in memory (when possible)
 - Typed
@@ -486,7 +486,7 @@ We also welcome contributions, please submit a PR.
 
 ### Development
 
-The library targets python 3.8 and higher. We use uv to manage the development environment.
+The library supports Python 3.8 through 3.14. We use uv to manage the development environment.
 
 Here is an example development workflow:
 

docs/features.md

@@ -14,7 +14,7 @@ Labelformat offers a robust set of features tailored to meet the diverse needs o
 
 - **Performance Optimizations:**
     - **Memory Conscious:** Processes datasets file-by-file to minimize memory usage.
-    - **Minimal Dependencies:** Targets Python 3.8 or higher, ensuring broad compatibility.
+    - **Minimal Dependencies:** Supports Python 3.8 through 3.14, ensuring broad compatibility.
 
 - **Cross-Platform Support:**
     - **Windows, Linux, and macOS:** Works seamlessly across all major operating systems.
@@ -59,4 +59,4 @@ Labelformat addresses the common challenges faced when dealing with diverse labe
 - **Scalability:** Handles large datasets with minimal memory footprint.
 - **Flexibility:** Supports a growing list of formats and tasks, adapting to evolving project needs.
 
-Explore our [Quick Start Guide](quick-start.md) to begin leveraging Labelformat's powerful features today!
+Explore our [Quick Start Guide](quick-start.md) to begin leveraging Labelformat's powerful features today!

docs/index.md

@@ -8,7 +8,7 @@
 
 ## Key Features
 - **Wide Format Support**: COCO, YOLO (v5-v12, v26), PascalVOC, KITTI, Labelbox, RT-DETR, RT-DETRv2, and more.
-- **Cross-Platform**: Compatible with Python 3.8+ on Windows, macOS, and Linux.
+- **Cross-Platform**: Compatible with Python 3.8 through 3.14 on Windows, macOS, and Linux.
 - **Flexible Usage**: Intuitive CLI and Python API.
 - **Efficient**: Memory-conscious, optimized for large datasets.
 - **Offline First**: Operates locally without data uploads.
@@ -66,4 +66,4 @@
 - [PyPI Package](https://pypi.org/project/labelformat/)
 - [Documentation](https://labelformat.com)
 
-Labelformat is maintained by [Lightly](https://www.lightly.ai).
+Labelformat is maintained by [Lightly](https://www.lightly.ai).

docs/installation.md

@@ -4,7 +4,7 @@ Installing **Labelformat** is straightforward. Follow the steps below to set up
 
 ## Prerequisites
 
-- **Python 3.8 or higher:** Ensure you have Python installed on Windows, Linux, or macOS.
+- **Python 3.8 through 3.14:** Ensure you have Python installed on Windows, Linux, or macOS.
 - **pip:** Python's package installer. It typically comes with Python installations.
 
 ## Installation using package managers

pyproject.toml

@@ -6,12 +6,22 @@ description = "A tool for converting computer vision label formats."
 readme = "README.md"
 license = { text = "MIT" }
 requires-python = ">=3.8"
+classifiers = [
+  "Programming Language :: Python :: 3",
+  "Programming Language :: Python :: 3.8",
+  "Programming Language :: Python :: 3.9",
+  "Programming Language :: Python :: 3.10",
+  "Programming Language :: Python :: 3.11",
+  "Programming Language :: Python :: 3.12",
+  "Programming Language :: Python :: 3.13",
+  "Programming Language :: Python :: 3.14",
+]
 dependencies = [
   "tqdm",
   "pyyaml",
   "pillow",
-  "pydantic==2.10.6",
-  "pydantic-xml==2.17.3",
+  "pydantic>=2.10.6",
+  "pydantic-xml>=2.17.3",
   "numpy",
   "fsspec>=2023.1.0",
 ]
@@ -32,7 +42,7 @@ dev = [
   "black",
   "isort",
   "flake8",
-  "pytest<8",
+  "pytest==8",
   "pytest-mock",
   "build",
   "twine",

src/labelformat/mask_utils.py

@@ -234,7 +234,7 @@ def _trace_outer_contour(binary_mask: NDArray[np.uint8]) -> List[Tuple[float, fl
     dirs = [(0, 1), (1, 1), (1, 0), (1, -1), (0, -1), (-1, -1), (-1, 0), (-1, 1)]
 
     def is_boundary(y: int, x: int) -> bool:
-        return 0 <= y < h and 0 <= x < w and boundary[y, x] == 1
+        return bool(0 <= y < h and 0 <= x < w and boundary[y, x] == 1)
 
     contour: List[Tuple[float, float]] = []
     y, x = sy, sx

src/labelformat/model/semantic_segmentation.py

@@ -43,7 +43,7 @@ def from_array(cls, array: NDArray[np.int_]) -> "SemanticSegmentationMask":
             else:
                 if cur_cat_id is not None:
                     category_id_rle.append((cur_cat_id, cur_run_length))
-                cur_cat_id = cat_id
+                cur_cat_id = int(cat_id)
                 cur_run_length = 1
         if cur_cat_id is not None:
             category_id_rle.append((cur_cat_id, cur_run_length))