ocr-projects-archive / ocr-tech /pyproject.toml
prosperious's picture
Add 1000 files (OCR_INDRA/.gitignore ...)
f9bc0c1 verified
Raw History Blame Contribute Delete
1.56 kB
[build-system]
requires = ["setuptools>=69", "wheel"]
build-backend = "setuptools.build_meta"
[project]
name = "ocr-tech"
version = "0.1.0"
description = "Hybrid Nepali + English OCR and document-structure extraction toolkit."
readme = "README.md"
requires-python = ">=3.11"
license = { text = "Apache-2.0" }
authors = [
{ name = "Ampixa" }
]
dependencies = []
[project.optional-dependencies]
datasets = [
"datasets>=3.0.0",
"h5py>=3.11.0",
"huggingface-hub>=0.25.0",
"pillow>=10.0.0"
]
eval = [
"pillow>=10.0.0",
"pymupdf>=1.24.0",
"rapidfuzz>=3.0.0",
"regex>=2024.0.0"
]
engines = [
"paddleocr>=3.0.0",
"pillow>=10.0.0"
]
surya = [
"surya-ocr>=0.20.0",
"pillow>=10.0.0"
]
hf-recognizer = [
"accelerate>=1.1.0",
"protobuf>=5.28.0",
"sentencepiece>=0.2.0",
"transformers>=4.57.0",
"torch>=2.5.0",
"pillow>=10.0.0"
]
vlm = [
"transformers>=4.50.0",
"torch>=2.5.0",
"pillow>=10.0.0"
]
dev = [
"pytest>=8.0.0"
]
# Everything the test suite imports; heavy engines (paddle, torch) stay optional and are skipped.
test = [
"pytest>=8.0.0",
"pillow>=10.0.0",
"pymupdf>=1.24.0",
"rapidfuzz>=3.0.0",
"regex>=2024.0.0",
"numpy>=1.26.0",
"opencv-python-headless>=4.8.0",
"fonttools>=4.50.0",
"uharfbuzz>=0.39.0",
"lxml>=5.0.0",
"pyyaml>=6.0.0",
"pyarrow>=15.0.0"
]
[project.scripts]
ocrtech = "ocrtech.cli:main"
[tool.setuptools.packages.find]
where = ["src"]
[tool.setuptools.package-data]
ocrtech = ["maps/*"]
[tool.pytest.ini_options]
pythonpath = ["src"]
testpaths = ["tests"]
addopts = "-q"