main.py

from dotenv import load_dotenv
load_dotenv()

import logging
logging.basicConfig(
    level=logging.INFO,
    format="%(asctime)s [%(levelname)s] %(name)s — %(message)s",
    datefmt="%H:%M:%S",
)

from fastapi import FastAPI
from fastapi.middleware.cors import CORSMiddleware
from contextlib import asynccontextmanager
from routers import yolo, clip, flow, pinecone
from services.yolo_service import YoloService
from services.clip_service import ClipService
from services.vision_service import VisionService
from services.pinecone_service import PineconeService

# Global service instances
yolo_service = YoloService()
clip_service = ClipService()
vision_service = VisionService()
pinecone_service = PineconeService()

@asynccontextmanager
async def lifespan(app: FastAPI):
    # Load models on startup
    print("Loading models...")
    await yolo_service.load_model()
    await clip_service.load_model()
    print("✅ All models ready")
    yield
    print("Shutting down...")

app = FastAPI(
    title="Fashion AI – ML Service",
    description="CLIP + YOLO inference service for fashion detection and vector generation",
    version="1.0.0",
    lifespan=lifespan,
)

app.add_middleware(
    CORSMiddleware,
    allow_origins=["*"],  # tighten to your Vercel domain in production if needed
    allow_credentials=False,
    allow_methods=["*"],
    allow_headers=["*"],
)

# Pass services to routers
app.state.yolo_service = yolo_service
app.state.clip_service = clip_service
app.state.vision_service = vision_service
app.state.pinecone_service = pinecone_service

app.include_router(yolo.router, prefix="/yolo", tags=["YOLO – Fashion Detection"])
app.include_router(clip.router, prefix="/clip", tags=["CLIP – Vector Generation"])
app.include_router(flow.router, prefix="/flow", tags=["Flow – End-to-End"])
app.include_router(pinecone.router, prefix="/pinecone", tags=["Pinecone"])

@app.get("/health")
def health():
    return { "status": "ok" }