backend/app/main.py

"""
DeepMed-AI Backend — main.py
FastAPI application entry point: app setup, lifespan, chat initialization.

Module layout:
  core/               — config, logging
  services/           — ChatService (HybridRAG), DatabaseService
  db/                 — SQLAlchemy session factory
  models/             — ORM models
  schemas/            — Pydantic request/response schemas
  api/v1/endpoints/   — health, chat, session route handlers
  api/v1/api.py       — router aggregator
  main.py             — FastAPI app + lifespan  ← you are here
"""

import asyncio
import os
import secrets
from contextlib import asynccontextmanager

# ── Fix SQLite for ChromaDB (MUST be before any chromadb import) ───────────────
# Python 3.10-slim ships SQLite 3.34, but ChromaDB requires >= 3.35.
# pysqlite3-binary provides a newer SQLite; we monkey-patch it in.
try:
    __import__("pysqlite3")
    import sys
    sys.modules["sqlite3"] = sys.modules.pop("pysqlite3")
except ImportError:
    pass  # Not needed on systems with SQLite >= 3.35 (e.g., Windows, newer Linux)

# ── Fix HuggingFace cache path (PHẢI đặt trước mọi import khác) ───────────────
# HF Spaces set HOME=/nonexistent → sentence-transformers/transformers crash.
# /tmp luôn writable trong Docker container trên HF Spaces.
os.environ.setdefault("HF_HOME", "/tmp/huggingface")
os.environ.setdefault("HUGGINGFACE_HUB_CACHE", "/tmp/huggingface/hub")
os.environ.setdefault("SENTENCE_TRANSFORMERS_HOME", "/tmp/sentence-transformers")
os.environ.setdefault("TRANSFORMERS_CACHE", "/tmp/transformers")
os.environ.setdefault("XDG_CACHE_HOME", "/tmp/.cache")

from fastapi import FastAPI, HTTPException
from fastapi.middleware.cors import CORSMiddleware
from fastapi.responses import FileResponse
from fastapi.staticfiles import StaticFiles
from starlette.middleware.sessions import SessionMiddleware

from app.api.v1.api import api_router
from app.core.config import CHAT_DB_PATH, CHROMA_DB_PATH, CHROMA_HF_DATASET, DATA_DIR
from app.core.logging_config import logger
from app.services.chat_service import chat_service
from app.services.database_service import db_service


# ── Lifespan ───────────────────────────────────────────────────────────────────
@asynccontextmanager
async def lifespan(app: FastAPI):
    """Application startup and shutdown lifecycle."""
    logger.info("Initializing DeepMed-AI System (TTYT Thanh Ba)...")

    # 1. Initialize database
    db_service.init_db()
    logger.info("Database initialized at %s", CHAT_DB_PATH)

    # 2. LFS pointer detection (early warning)
    if os.path.exists(DATA_DIR):
        lfs_pointers = []
        for root, _dirs, files in os.walk(DATA_DIR):
            for fname in files:
                if fname.endswith((".pdf", ".docx", ".xlsx")):
                    fp = os.path.join(root, fname)
                    if os.path.getsize(fp) < 500:
                        lfs_pointers.append(fp)
        if lfs_pointers:
            logger.error(
                "LFS POINTER STUBS DETECTED — %d data file(s) are Git LFS pointers, "
                "not real content. RAG will be EMPTY!\n"
                "Fix: run `git lfs pull` in the project root, then rebuild the Docker image.\n"
                "Affected files:\n  %s",
                len(lfs_pointers),
                "\n  ".join(lfs_pointers),
            )
    else:
        logger.warning("DATA directory not found at %s — RAG will have no data", DATA_DIR)

    # 3. Download ChromaDB from HuggingFace Dataset if not present locally
    if not os.path.exists(CHROMA_DB_PATH) or not os.listdir(CHROMA_DB_PATH):
        logger.info("ChromaDB not found locally, downloading from HF Dataset: %s", CHROMA_HF_DATASET)
        try:
            from huggingface_hub import snapshot_download
            snapshot_download(
                repo_id=CHROMA_HF_DATASET,
                repo_type="dataset",
                local_dir=CHROMA_DB_PATH,
                local_dir_use_symlinks=False,
            )
            logger.info("ChromaDB downloaded successfully to %s", CHROMA_DB_PATH)
        except Exception as e:
            logger.error("Failed to download ChromaDB from HF Dataset: %s", e)
    else:
        logger.info("ChromaDB already exists at %s", CHROMA_DB_PATH)

    # 4. Initialize HybridRAG bot (ChromaDB + BM25 + Reranker)
    loop = asyncio.get_event_loop()
    await loop.run_in_executor(None, chat_service.initialize)
    logger.info("DeepMed-AI System Ready!")

    yield

    logger.info("Shutting down DeepMed-AI...")


# ── FastAPI App ────────────────────────────────────────────────────────────────
app = FastAPI(
    title="DeepMed-AI API",
    description="Trí tuệ nhân tạo của TTYT khu vực Thanh Ba",
    version="3.0.0",
    lifespan=lifespan,
)

# ── CORS ────────────────────────────────────────────────────────────────────────
_raw_origins = os.getenv(
    "ALLOWED_ORIGINS",
    "http://localhost:5173,http://localhost:3000,http://localhost:7860",
)
_allowed_origins = [o.strip() for o in _raw_origins.split(",") if o.strip()]

app.add_middleware(
    CORSMiddleware,
    allow_origins=_allowed_origins,
    allow_credentials=True,
    allow_methods=["GET", "POST", "DELETE", "OPTIONS"],
    allow_headers=["Content-Type", "Authorization", "X-Requested-With"],
)

# ── Session ────────────────────────────────────────────────────────────────────
_session_secret = os.getenv("SESSION_SECRET", secrets.token_hex(32))
app.add_middleware(SessionMiddleware, secret_key=_session_secret)

# ── API Routes ─────────────────────────────────────────────────────────────────
app.include_router(api_router)

# ── Serve React Frontend (SPA) ─────────────────────────────────────────────────
STATIC_DIR = os.path.normpath(
    os.path.join(os.path.dirname(__file__), "..", "static")
)

if os.path.isdir(STATIC_DIR):
    assets_dir = os.path.join(STATIC_DIR, "assets")
    if os.path.isdir(assets_dir):
        app.mount("/assets", StaticFiles(directory=assets_dir), name="assets")

    @app.get("/{path:path}")
    async def serve_static(path: str):
        # Never intercept API routes — let 404 bubble up naturally
        if path.startswith("api/"):
            raise HTTPException(status_code=404, detail="Not found")
        file_path = os.path.join(STATIC_DIR, path)
        if os.path.isfile(file_path):
            return FileResponse(file_path)
        # SPA fallback: return index.html for client-side routing
        return FileResponse(os.path.join(STATIC_DIR, "index.html"))

else:
    logger.warning("Static directory not found at %s. Frontend will not be served.", STATIC_DIR)


# ── Entry Point ────────────────────────────────────────────────────────────────
if __name__ == "__main__":
    import uvicorn

    uvicorn.run(app, host="0.0.0.0", port=8000)