main.py

"""
FastAPI application entry point — Detector de Texto Generado por IA
"""
import os
from contextlib import asynccontextmanager

from fastapi import FastAPI, HTTPException
from fastapi.middleware.cors import CORSMiddleware
from fastapi.staticfiles import StaticFiles

from schemas      import PredictRequest, PredictResponse
from model_loader import load_model, predict as run_inference


# ── Lifespan: eager model loading on startup ───────────────────────────────────
@asynccontextmanager
async def lifespan(app: FastAPI):
    load_model()
    yield


# ── App ────────────────────────────────────────────────────────────────────────
app = FastAPI(
    title       = "Synthetic Sentinel API",
    description = "Detección de texto generado por IA — Universidad de Guayaquil",
    version     = "2.0.0",
    lifespan    = lifespan,
)

app.add_middleware(
    CORSMiddleware,
    allow_origins  = ["*"],
    allow_methods  = ["GET", "POST"],
    allow_headers  = ["*"],
)


# ── Routes ─────────────────────────────────────────────────────────────────────
@app.get("/health", summary="Health check")
def health():
    return {"status": "ok"}


@app.post("/predict", response_model=PredictResponse, summary="Analizar texto")
def predict(body: PredictRequest):
    """Receive Spanish text and return AI-detection probability."""
    if not body.text.strip():
        raise HTTPException(status_code=422, detail="El texto no puede estar vacío.")

    try:
        result = run_inference(body.text)
    except Exception as exc:
        raise HTTPException(status_code=500, detail=str(exc))

    return PredictResponse(**result)


# ── Static frontend (production) ───────────────────────────────────────────────
_static_dir = os.path.join(os.path.dirname(__file__), "..", "frontend", "dist")
if os.path.isdir(_static_dir):
    app.mount("/", StaticFiles(directory=_static_dir, html=True), name="frontend")