import os
import requests
import json
import time
import re
from dotenv import load_dotenv

# Charger .env.global
load_dotenv("../.env.global")
AIANALASYS_APIKEY = os.getenv("AIANALASYS_APIKEY")

def get_mammouth_models():
    url = "https://api.mammouth.ai/public/models"
    try:
        # Désactiver les warnings InsecureRequest car verify=False est utilisé
        requests.packages.urllib3.disable_warnings()
        response = requests.get(url, verify=False)
        response.raise_for_status()
        return response.json().get('data', [])
    except Exception as e:
        print(f"Error fetching Mammouth models: {e}")
        return []

def get_aa_data():
    url = "https://artificialanalysis.ai/api/v2/data/llms/models"
    headers = {"x-api-key": AIANALASYS_APIKEY}
    try:
        response = requests.get(url, headers=headers)
        response.raise_for_status()
        return response.json().get('data', [])
    except Exception as e:
        print(f"Error fetching AA data: {e}")
        return []

def clean_id(model_id):
    # Nettoyage agressif pour favoriser le mapping
    id_clean = re.sub(r'-\d{4,8}', '', model_id.lower())
    id_clean = id_clean.replace('-latest', '').replace('-preview', '').replace('-instruct', '')
    return id_clean.strip()

def generate_markdown(models_data):
    categories = {}
    for m in models_data:
        cat = m.get('category', 'General')
        if cat not in categories: categories[cat] = []
        categories[cat].append(m)
        
    md = "# Table des Modèles Mammouth.ai\n\n"
    md += "*Généré automatiquement à partir des benchmarks d'Artificial Analysis et des tarifs Mammouth.ai.*\n\n"
    md += f"Dernière mise à jour : {time.strftime('%Y-%m-%d %H:%M:%S')}\n\n"
    
    order = ['Coding', 'Agents', 'General']
    sorted_cats = sorted(categories.keys(), key=lambda x: order.index(x) if x in order else 99)

    for cat in sorted_cats:
        md += f"## {cat}\n\n"
        md += "| Modèle | Prix (In / Out / 1M) | Score (Intelligence) | Vitesse (TPS) |\n"
        md += "| :--- | :--- | :--- | :--- |\n"
        models = categories[cat]
        # Tri : Score (desc), puis Nom
        models.sort(key=lambda x: (x.get('score') or 0), reverse=True)
        for m in models:
            score_str = f"**{m['score']:.1f}**" if m['score'] else "N/A"
            speed_str = f"{m['speed']:.1f}" if m['speed'] else "N/A"
            md += f"| {m['name']} | ${m['price_in']:.2f} / ${m['price_out']:.2f} | {score_str} | {speed_str} |\n"
        md += "\n"
    return md

def main():
    print("Fetching data from Mammouth and Artificial Analysis...")
    m_models = get_mammouth_models()
    aa_data = get_aa_data()
    
    # Mapping table (slug -> data)
    aa_map = {}
    for aa_m in aa_data:
        slug = aa_m.get('slug', '').lower()
        name = aa_m.get('name', '').lower()
        if slug: aa_map[slug] = aa_m
        if name: aa_map[name] = aa_m

    enriched = []
    for m in m_models:
        m_id = m.get('id', '')
        info = m.get('model_info', {})
        if not m_id: continue

        m_id_clean = clean_id(m_id)
        short_id = m_id_clean.split('/')[-1]
        
        # Match mapping
        aa_info = aa_map.get(m_id_clean) or aa_map.get(short_id)
        
        # Recherche floue (ex: claude-3-5-sonnet -> claude-3.5-sonnet)
        if not aa_info:
            normalized_m_id = m_id_clean.replace('-', '').replace('.', '')
            for key, val in aa_map.items():
                if key.replace('-', '').replace('.', '') == normalized_m_id:
                    aa_info = val
                    break

        price_in = float(info.get('input_cost_per_token', 0)) * 1000000
        price_out = float(info.get('output_cost_per_token', 0)) * 1000000
        
        category = "General"
        if any(x in m_id_clean for x in ['coding', 'code', 'starcoder', 'codestral', 'coder']):
            category = "Coding"
        elif any(x in m_id_clean for x in ['agent', 'hermes', 'tool', 'function', 'sonar']):
            category = "Agents"

        score = None
        speed = None
        if aa_info:
            evals = aa_info.get('evaluations', {})
            # On prend le score coding si c'est la catégorie, sinon intelligence index
            score = evals.get('artificial_analysis_coding_index') if category == "Coding" else None
            if not score:
                score = evals.get('artificial_analysis_intelligence_index')
            
            speed = aa_info.get('median_output_tokens_per_second')

        enriched.append({
            'name': m_id,
            'price_in': price_in,
            'price_out': price_out,
            'score': score,
            'speed': speed,
            'category': category
        })

    # On ne garde que les modèles avec prix > 0
    final = [x for x in enriched if x['price_in'] > 0]
    
    with open("README.md", "w", encoding="utf-8") as f:
        f.write(generate_markdown(final))
    
    print(f"Success! {len(final)} models processed.")

if __name__ == "__main__":
    main()