AIRegulation-DocAnalysis/backend/requirements.txt

# ── Web framework ─────────────────────────────────────────────────────────────
fastapi>=0.110.0
uvicorn[standard]>=0.27.0
python-multipart>=0.0.9
# MCP server module (backend/app/mcp/) — pin >=2.0.0: that release renamed the
# older "FastMCP" class to "MCPServer" (mcp.server.MCPServer), which is the
# class actually used in app/mcp/server.py.
mcp>=2.0.0

# ── Config & utilities ────────────────────────────────────────────────────────
pydantic>=2.0.0
pydantic-settings>=2.0.0
python-dotenv>=1.0.0
loguru>=0.7.0
httpx>=0.25.0
beautifulsoup4>=4.12.0
lxml>=5.0.0
tiktoken>=0.5.0
tenacity>=8.2.0
# Regulatory signal crawling (backend/app/infrastructure/perception/) — main-content
# extraction from crawled regulation detail pages and character-level diff for change
# detection. Import name for diff-match-patch is diff_match_patch (underscored).
trafilatura>=2.0.0
diff-match-patch>=20241021

# ── Auth ──────────────────────────────────────────────────────────────────────
python-jose[cryptography]>=3.3.0
# passlib is incompatible with bcrypt>=4.0 (removed __about__, strict 72-byte limit).
# Pin bcrypt to 3.x until passlib ships a fix.
passlib[bcrypt]>=1.7.4
bcrypt>=3.2.0,<4.0.0

# ── Async task queue ──────────────────────────────────────────────────────────
celery>=5.3.0
redis>=4.5.0

# ── Storage & databases ───────────────────────────────────────────────────────
pymilvus>=2.4.0
minio>=7.1.0
psycopg2-binary>=2.9.0

# ── Document parsing ─────────────────────────────────────────────────────────
pymupdf>=1.24.0
python-docx>=1.1.0
alibabacloud-docmind-api20220711>=1.0.6
alibabacloud-tea-openapi>=0.3.11
alibabacloud-tea-util>=0.3.13

# ── RAG / LangChain ───────────────────────────────────────────────────────────
langchain>=0.1.0
langchain-milvus>=0.1.0
numpy>=1.24.0

# ── Testing ───────────────────────────────────────────────────────────────────
pytest>=7.4.0
pytest-asyncio>=0.21.0
fakeredis>=2.0.0