"""
Logique ML pour les recommandations d'outfits
Adapté de votre fichier recommendations.py
"""

import numpy as np
import requests
from sklearn.metrics.pairwise import cosine_similarity
import json

# Configuration météo
OPENWEATHER_API_KEY = "a92f907ace22631f8af40374ae0b30b6"

def get_weather(city: str):
    """Récupère la météo depuis OpenWeather API"""
    try:
        url = f"https://api.openweathermap.org/data/2.5/weather?q={city}&appid={OPENWEATHER_API_KEY}&units=metric"
        response = requests.get(url, timeout=5)
        data = response.json()
        return {
            "temperature": data["main"]["temp"],
            "condition": data["weather"][0]["main"]
        }
    except:
        return {"temperature": 20, "condition": "Clear"}

def get_season_from_weather(temp: float):
    """Détermine la saison selon la température"""
    if temp > 25:
        return "summer"
    elif temp > 17:
        return "spring"
    elif temp > 0:
        return "fall"
    return "winter"

def recommend_outfit_ml(clothes_data: list, preference: str, city: str):
    """
    Recommande un outfit complet basé sur ML
    
    Args:
        clothes_data: Liste des vêtements disponibles
        preference: Style préféré
        city: Ville pour la météo
    
    Returns:
        Dict avec l'outfit recommandé et l'explication
    """
    # 1. Obtenir la météo
    weather = get_weather(city)
    season = get_season_from_weather(weather["temperature"])
    
    # 2. Filtrer par style et saison
    pref_lower = preference.lower()
    
    def matches_season(item_season, target_season):
        if not item_season or item_season.lower() in ["all", "toutes"]:
            return True
        return item_season.lower() == target_season
    
    tops = [c for c in clothes_data if c.get("category") == "top" and c.get("style") == pref_lower and matches_season(c.get("season"), season)]
    bottoms = [c for c in clothes_data if c.get("category") == "bottom" and c.get("style") == pref_lower and matches_season(c.get("season"), season)]
    footwear = [c for c in clothes_data if c.get("category") in ["footwear", "shoes"] and c.get("style") == pref_lower and matches_season(c.get("season"), season)]
    
    # 3. Vérifier si on a assez de vêtements
    if not tops or not bottoms or not footwear:
        return {
            "success": False,
            "message": f"Pas assez de vêtements '{preference}' pour la saison '{season}'",
            "weather": weather,
            "season": season
        }
    
    # 4. Sélectionner le meilleur outfit (basé sur les scores)
    top = max(tops, key=lambda x: x.get("score", 0))
    bottom = max(bottoms, key=lambda x: x.get("score", 0))
    shoe = max(footwear, key=lambda x: x.get("score", 0))
    
    return {
        "success": True,
        "outfit": {
            "top": top["id"],
            "bottom": bottom["id"],
            "footwear": shoe["id"]
        },
        "explanation": {
            "top": {
                "reason": f"Best rated (Score: {top.get('score', 0):.2f})",
                "score": top.get("score", 0)
            },
            "bottom": {
                "reason": f"Best match (Score: {bottom.get('score', 0):.2f})",
                "score": bottom.get("score", 0)
            },
            "footwear": {
                "reason": f"Best match (Score: {shoe.get('score', 0):.2f})",
                "score": shoe.get("score", 0)
            }
        },
        "weather": weather,
        "season": season,
        "preference": preference
    }