36 lines
579 B
Plaintext
36 lines
579 B
Plaintext
# Azure services
|
|
azure-ai-formrecognizer==3.3.0
|
|
azure-core>=1.29.0
|
|
openai>=1.0.0
|
|
|
|
# PDF processing
|
|
pdf2image==1.17.0
|
|
Pillow>=10.0.0
|
|
pytesseract==0.3.13 # Fallback OCR
|
|
pdfplumber==0.10.3 # Enhanced table extraction
|
|
PyPDF2>=3.0.1
|
|
|
|
# CLI and utilities
|
|
click==8.1.7
|
|
pathlib2==2.3.7.post1
|
|
|
|
# Data handling
|
|
numpy>=1.24.0
|
|
python-dateutil>=2.8.2
|
|
|
|
# For better logging
|
|
colorlog>=6.7.0
|
|
|
|
# Development tools (optional)
|
|
pytest>=7.4.0
|
|
black>=23.0.0
|
|
flake8>=6.0.0
|
|
|
|
# Type hints
|
|
typing-extensions>=4.0.0
|
|
dataclasses>=0.6
|
|
|
|
# For text processing
|
|
nltk>=3.8.1
|
|
textstat>=0.7.3
|
|
python-dotenv==1.0.0 |