"""
FastAPI service exposing BinhQuocNguyen/food‑recognition‑model.
"""

# ------------------------------------------------------------
# 1️⃣ Imports
# ------------------------------------------------------------
from fastapi import FastAPI, HTTPException
from pydantic import BaseModel
import base64
import io

from PIL import Image
from transformers import pipeline

# ------------------------------------------------------------
# 2️⃣ Load the model once at startup (fast, no re‑load per request)
# ------------------------------------------------------------
classifier = pipeline(
    "image-classification",
    model="BinhQuocNguyen/food-recognition-model"
)

# ------------------------------------------------------------
# 3️⃣ Simple in‑memory “nutritional DB”
#    (calories per 100 g, average portion weight in grams)
# ------------------------------------------------------------
nutrient_db = {
    "Apple":  {"calories_per_100g": 52,  "portion_g": 182},
    "Banana": {"calories_per_100g": 89,  "portion_g": 118},
    "Orange": {"calories_per_100g": 43,  "portion_g": 131},
    "Pizza":  {"calories_per_100g": 266, "portion_g": 200},
    "Bread":  {"calories_per_100g": 265, "portion_g": 30},
    # ← add the rest of your favourite foods here
}

# ------------------------------------------------------------
# 4️⃣ Pydantic model for the incoming JSON payload
# ------------------------------------------------------------
class ImageRequest(BaseModel):
    """Base64‑encoded image sent by the client."""
    image: str


# ------------------------------------------------------------
# 5️⃣ FastAPI app & health‑check endpoint
# ------------------------------------------------------------
app = FastAPI()


@app.get("/")
def root():
    """Simple health‑check."""
    return {"message": "Food‑Recognition API is up"}


# ------------------------------------------------------------
# 6️⃣ Main inference endpoint
# ------------------------------------------------------------
@app.post("/analyze")
def analyze(request: ImageRequest):
    """
    1️⃣ Decode the base64 image.
    2️⃣ Run the classifier.
    3️⃣ Look up (or fall back to) nutritional information.
    4️⃣ Return a JSON response.
    """
    # ---- 1️⃣ decode the base64 image ---------------------------------
    try:
        raw_bytes = base64.b64decode(request.image)
        pil_img = Image.open(io.BytesIO(raw_bytes)).convert("RGB")
    except Exception as e:  # pragma: no cover
        raise HTTPException(status_code=400, detail="Invalid base64 image") from e

    # ---- 2️⃣ run the classifier --------------------------------------
    results = classifier(pil_img)
    if not results:
        raise HTTPException(status_code=500, detail="Model returned no results")

    top = results[0]
    label = top.get("label")
    confidence = top.get("score")

    # ---- 3️⃣ look up nutrition data ----------------------------------
    # Fallback values if the label isn’t in the DB
    nutrition = nutrient_db.get(label, {"calories_per_100g": 0, "portion_g": 100})
    calories_per_100g = nutrition["calories_per_100g"]
    portion_g = nutrition["portion_g"]

    # ---- 4️⃣ estimated calories for the default portion size ----------
    est_calories = calories_per_100g * (portion_g / 100.0)

    # ---- 5️⃣ build the JSON response --------------------------------
    return {
        "label": label,
        "confidence": confidence,
        "estimated_portion_g": portion_g,
        "calories_per_100g": calories_per_100g,
        "estimated_calories": round(est_calories, 2),
    }