pyproject.toml

[build-system]
requires = ["setuptools>=68"]
build-backend = "setuptools.build_meta"

[project]
name = "nas_file_organizer"
version = "0.4.3"
description = "Content-aware file sorter with OCR for NAS"
authors = [{name = "Tamas Kiss", email = "thomas.kaen.freelance@gmail.com"}]
readme = "README.md"
requires-python = ">=3.10"
dependencies = [
  # --- Core app / web ---
  "fastapi>=0.115",
  "uvicorn>=0.30",
  "jinja2>=3.1",
  "python-multipart>=0.0.9",
  "rich>=13.7",

  # --- Rules / YAML ---
  "ruyaml>=0.91",

  # --- File parsing & OCR (optional at runtime unless you enable OCR) ---
  "pymupdf>=1.24",        # PDF text (fast + reliable)
  "pytesseract>=0.3.10",  # OCR bindings (requires Tesseract binary installed)
  "Pillow>=10.0",         # imaging for OCR
  "python-docx>=1.1",     # DOCX text
  "openpyxl>=3.1",        # XLSX text
  "chardet>=5.2",         # robust TXT decode

  # --- ML (training & inference) ---
  "scikit-learn>=1.4",
  "numpy>=1.26",
  "scipy>=1.10",          # sklearn runtime dep; pin to avoid resolver pain
  "joblib>=1.3",          # sklearn model persistence

  # --- Fuzzy / matching ---
  "rapidfuzz>=3.9",

  # --- Scheduling (auto weekly retrain) ---
  "apscheduler>=3.10",

  # --- Optional but useful ---
  "python-dateutil>=2.9", # APScheduler & timestamp parsing niceties
  "watchdog>=4.0",        # (later) folder watcher
]

[project.optional-dependencies]
ml = ["scikit-learn>=1.4", "numpy>=1.26", "scipy>=1.10", "joblib>=1.3"]
ocr = ["pymupdf>=1.24", "pytesseract>=0.3.10", "Pillow>=10.0"]
docs = ["rich>=13.7"]
watch = ["watchdog>=4.0"]

dev = [
  "pytest",
  "ruff",
  "mypy",
]

[tool.setuptools]
packages = ["nas_file_organizer"]

[project.scripts]
nas-organize = "nas_file_organizer.adapters.cli:main"
nas-watch    = "nas_file_organizer.adapters.watch:main"
nas-web      = "nas_file_organizer.adapters.web:main"
nas-train    = "nas_file_organizer.adapters.train_cli:main"