tasks/worker_celery.py

import os
import json
from celery import Celery
from crewai import Crew, Process
from dotenv import load_dotenv

# Charger les variables d'environnement depuis un fichier .env
load_dotenv()

# --- Import des modules de votre application ---
from src.deep_learning_analyzer import MultiModelInterviewAnalyzer
from src.rag_handler import RAGHandler
from src.crew.agents import report_generator_agent
from src.crew.tasks import generate_report_task

# --- Configuration de Celery avec Upstash ---
# On récupère les URLs et le token depuis les variables d'environnement
UPSTASH_REDIS_URL = os.environ.get("UPSTASH_REDIS_URL")
UPSTASH_REDIS_TOKEN = os.environ.get("UPSTASH_REDIS_TOKEN")

# On formate l'URL pour Celery en incluant le token
# Format: rediss://:<token>@<url_sans_https>
broker_url = f"rediss://:{UPSTASH_REDIS_TOKEN}@{UPSTASH_REDIS_URL.replace('https://', '')}"

celery_app = Celery(
    'worker_celery',
    broker=broker_url,
    backend=broker_url,
    broker_connection_retry_on_startup=True,
    include=['tasks.worker_celery']
)

celery_app.conf.update(
    task_serializer='json',
    result_serializer='json',
    accept_content=['json'],
    timezone='Europe/Paris',
    enable_utc=True,
    # Configuration SSL nécessaire pour Upstash
    broker_use_ssl={
        'ssl_cert_reqs': 'CERT_NONE'
    },
    redis_backend_use_ssl={
        'ssl_cert_reqs': 'CERT_NONE'
    }
)

# --- La tâche reste la même ---
@celery_app.task(name="tasks.run_interview_analysis")
def run_interview_analysis_task(conversation_history: list, job_description_text: list):
    """
    Tâche Celery qui exécute l'analyse complète de l'entretien en arrière-plan.
    """
    print("Début de l'analyse via le worker Celery...")
    
    # Étape 1: Analyse DL
    analyzer = MultiModelInterviewAnalyzer()
    structured_analysis = analyzer.run_full_analysis(conversation_history, job_description_text)

    # Étape 2: Enrichissement RAG
    rag_handler = RAGHandler()
    rag_feedback = []
    if structured_analysis.get("intent_analysis"):
        for intent in structured_analysis["intent_analysis"]:
            query = f"Conseils pour un candidat qui cherche à {intent['labels'][0]}"
            rag_feedback.extend(rag_handler.get_relevant_feedback(query))
    
    if structured_analysis.get("sentiment_analysis"):
        for sentiment_group in structured_analysis["sentiment_analysis"]:
            for sentiment in sentiment_group:
                if sentiment['label'] == 'stress' and sentiment['score'] > 0.6:
                    rag_feedback.extend(rag_handler.get_relevant_feedback("gestion du stress en entretien"))

    unique_feedback = list(set(rag_feedback))

    # Étape 3: Génération du rapport avec CrewAI
    interview_crew = Crew(
        agents=[report_generator_agent],
        tasks=[generate_report_task],
        process=Process.sequential,
        verbose=False,
        telemetry=False
    )

    final_report = interview_crew.kickoff(inputs={
        'structured_analysis_data': json.dumps(structured_analysis, indent=2, ensure_ascii=False),
        'rag_contextual_feedback': "\n- ".join(unique_feedback)
    })
    
    print("Rapport final généré. Tâche terminée.")
    return final_report