circuitforge-core/pyproject.toml
pyr0ball bbb146b361 feat(documents): add PDFExtractor text-layer extraction and PageChunk
Adds circuitforge_core/documents/pdf.py with:
- PageChunk frozen dataclass (page_number, text, source, word_count)
- PDFExtractor.chunk_pages() — pdfplumber text-layer per page, OCR fallback via pytesseract for sparse pages
- Module-level graceful ImportError guard on pdfplumber (patchable, follows cf-core optional-extra pattern)
- pdf and pdf-ocr optional extras declared in pyproject.toml

3 tests, all passing.
2026-05-04 08:33:10 -07:00

140 lines
2.7 KiB
TOML

[build-system]
requires = ["setuptools>=68"]
build-backend = "setuptools.build_meta"
[project]
name = "circuitforge-core"
version = "0.18.0"
description = "Shared scaffold for CircuitForge products (MIT)"
requires-python = ">=3.11"
dependencies = [
"pyyaml>=6.0",
"requests>=2.31",
"openai>=1.0",
]
[project.optional-dependencies]
community = [
"psycopg2>=2.9",
]
manage = [
"platformdirs>=4.0",
"typer[all]>=0.12",
]
text-llamacpp = [
"llama-cpp-python>=0.2.0",
]
text-transformers = [
"torch>=2.0",
"transformers>=4.40",
"accelerate>=0.27",
]
text-transformers-4bit = [
"circuitforge-core[text-transformers]",
"bitsandbytes>=0.43",
]
stt-faster-whisper = [
"faster-whisper>=1.0",
]
stt-service = [
"circuitforge-core[stt-faster-whisper]",
"fastapi>=0.110",
"uvicorn[standard]>=0.29",
"python-multipart>=0.0.9",
]
tts-chatterbox = [
"chatterbox-tts>=0.1",
"torchaudio>=2.0",
]
tts-service = [
"circuitforge-core[tts-chatterbox]",
"fastapi>=0.110",
"uvicorn[standard]>=0.29",
"python-multipart>=0.0.9",
]
musicgen-audiocraft = [
"audiocraft>=1.3.0",
"torch>=2.0",
"torchaudio>=2.0",
"einops>=0.7",
"flashy>=0.0.1",
"num2words>=0.5",
]
musicgen-service = [
"circuitforge-core[musicgen-audiocraft]",
"fastapi>=0.110",
"uvicorn[standard]>=0.29",
"python-multipart>=0.0.9",
]
vision-siglip = [
"torch>=2.0",
"transformers>=4.40",
"Pillow>=10.0",
]
vision-vlm = [
"torch>=2.0",
"transformers>=4.40",
"Pillow>=10.0",
"accelerate>=0.27",
]
vision-service = [
"circuitforge-core[vision-siglip]",
"fastapi>=0.110",
"uvicorn[standard]>=0.29",
"python-multipart>=0.0.9",
]
reranker-bge = [
"FlagEmbedding>=1.2",
]
reranker-qwen3 = [
"torch>=2.0",
"transformers>=4.40",
"accelerate>=0.27",
]
reranker-cross-encoder = [
"sentence-transformers>=3.0",
]
reranker-cohere = [
"cohere>=5.0",
]
reranker-service = [
"circuitforge-core[reranker-qwen3]",
"fastapi>=0.110",
"uvicorn[standard]>=0.29",
]
gestures-mediapipe = [
"mediapipe>=0.10",
"opencv-python>=4.8",
"numpy>=1.24",
]
pdf = [
"pdfplumber>=0.11",
]
pdf-ocr = [
"circuitforge-core[pdf]",
"pytesseract>=0.3",
"Pillow>=10.0",
]
dev = [
"circuitforge-core[manage]",
"pytest>=8.0",
"pytest-asyncio>=0.23",
"fastapi>=0.110",
"httpx>=0.27",
"ruff>=0.4",
"mypy>=1.10",
]
[project.scripts]
cf-manage = "circuitforge_core.manage.cli:app"
[tool.setuptools.packages.find]
where = ["."]
include = ["circuitforge_core*"]
[tool.setuptools.package-data]
"circuitforge_core.community.migrations" = ["*.sql"]
[tool.pytest.ini_options]
testpaths = ["tests"]
asyncio_mode = "auto"