requirements.txt

# ePubs PDF Accessibility Enhancement - Consolidated Requirements
# Production-ready simplified system (Stage 2 & 3 complete)
# Last updated: 2025-10-29

# ==============================================================================
# PRODUCTION DEPENDENCIES
# ==============================================================================

# Core PDF Processing
PyMuPDF>=1.23.0,<1.25.0       # PDF manipulation and text extraction
pikepdf>=8.3.0,<9.0.0          # PDF structure modification and enhancement

# Image Processing
Pillow>=10.0.0,<11.0.0         # Image manipulation for OCR
numpy>=1.24.0,<2.0.0           # Numerical operations for image processing

# OCR
ocrmypdf>=16.0.0,<17.0.0       # Integrated OCR solution with proper text positioning
pytesseract>=0.3.10,<1.0.0     # Tesseract OCR interface (used by ocrmypdf)

# Progress Tracking
tqdm>=4.66.0,<5.0.0            # Progress bars for batch processing

# Configuration
pyyaml>=6.0.0,<7.0.0           # YAML configuration files
python-decouple>=3.6,<4.0      # Environment configuration

# CLI Interface
typer>=0.12.0,<1.0.0           # CLI framework
rich>=13.0.0,<14.0.0           # Rich terminal output

# Utilities
python-magic>=0.4.27,<1.0.0    # File type detection

# ==============================================================================
# DEVELOPMENT DEPENDENCIES (Optional - for contributors)
# ==============================================================================
# Uncomment if you're developing/contributing to the project:

# pytest>=7.0.0,<8.0.0          # Testing framework
# pytest-cov>=4.0.0,<5.0.0      # Test coverage reporting
# black>=23.0.0,<24.0.0         # Code formatting
# isort>=5.12.0,<6.0.0          # Import sorting

# ==============================================================================
# SYSTEM REQUIREMENTS
# ==============================================================================
# Tesseract OCR must be installed separately:
# - macOS:   brew install tesseract
# - Ubuntu:  sudo apt-get install tesseract-ocr
# - Windows: Download from https://github.com/UB-Mannheim/tesseract/wiki

# ==============================================================================
# INSTALLATION
# ==============================================================================
# pip install -r requirements.txt

# ==============================================================================
# NOTES
# ==============================================================================
# This simplified requirements file replaces all previous variations:
# - requirements.txt (old Phase 2B with 33+ packages)
# - requirements-simple.txt (Stage 2 minimal deps)
# - requirements-dev.txt (development tools)
# - requirements-lock.txt (pinned versions)
# - requirements-docs.txt (document format support)
#
# Total production dependencies: 12 packages (down from 33+)
# This is the ONLY requirements file needed for the project.