# Web Framework
streamlit>=1.28.0

# AI and LLM
langchain-core>=0.3.0
langchain-community>=0.3.0
langchain-openai>=0.3.0
langgraph>=0.4.0
openai>=1.78.0

# Document Processing (LangChain)
unstructured>=0.12.0
unstructured[pdf]>=0.12.0
unstructured[docx]>=0.12.0
unstructured[md]>=0.12.0
python-magic>=0.4.27

# Web Scraping
selenium>=4.15.0
beautifulsoup4>=4.12.0
requests>=2.32.0
webdriver-manager>=4.0.0

# Document Processing (Fallback)
PyPDF2>=3.0.0
pdfplumber>=0.10.0
python-docx>=1.1.0
python-docx2txt>=0.8
markdown>=3.5.0
pypdf>=5.0.0

# Data Processing
pydantic>=2.11.0
PyYAML>=6.0.0
python-dotenv>=1.1.0
pandas>=2.0.0
numpy>=1.24.0

# Utilities
tqdm>=4.67.0
tenacity>=9.1.0
xxhash>=3.5.0

# File type detection
filetype>=1.2.0

# 可选依赖（Token估算）
# tiktoken>=0.5.0

# 可选依赖（OpenAI API支持）
# openai>=1.0.0 