Nicolas Cantu
f50481cc38
Some checks failed
publish-images / docker-build-and-push (push) Failing after 23s
83 lines
2.1 KiB
Python
83 lines
2.1 KiB
Python
"""
|
|
Tâches OCR pour le traitement des documents
|
|
"""
|
|
import logging
|
|
from typing import Dict, Any
|
|
from celery_app import app
|
|
|
|
logger = logging.getLogger(__name__)
|
|
|
|
@app.task(bind=True, name='ocr.process_document')
|
|
def process_document_ocr(self, doc_id: str, context: Dict[str, Any]) -> Dict[str, Any]:
|
|
"""
|
|
Traitement OCR d'un document
|
|
|
|
Args:
|
|
doc_id: Identifiant du document
|
|
context: Contexte de traitement
|
|
|
|
Returns:
|
|
Résultat de l'OCR
|
|
"""
|
|
try:
|
|
logger.info(f"Début de l'OCR pour le document {doc_id}")
|
|
|
|
# Mise à jour du statut
|
|
self.update_state(
|
|
state='PROGRESS',
|
|
meta={'current_step': 'ocr_processing', 'progress': 0}
|
|
)
|
|
|
|
# TODO: Implémenter le traitement OCR réel
|
|
# Pour l'instant, simulation
|
|
import time
|
|
time.sleep(2) # Simulation du traitement
|
|
|
|
result = {
|
|
'doc_id': doc_id,
|
|
'status': 'completed',
|
|
'text_extracted': 'Texte extrait simulé',
|
|
'confidence': 0.95,
|
|
'pages_processed': 1,
|
|
'processing_time': 2.0
|
|
}
|
|
|
|
logger.info(f"OCR terminé pour le document {doc_id}")
|
|
return result
|
|
|
|
except Exception as e:
|
|
logger.error(f"Erreur lors de l'OCR du document {doc_id}: {e}")
|
|
raise
|
|
|
|
@app.task(name='ocr.batch_process')
|
|
def batch_process_ocr(doc_ids: list) -> Dict[str, Any]:
|
|
"""
|
|
Traitement OCR en lot
|
|
|
|
Args:
|
|
doc_ids: Liste des identifiants de documents
|
|
|
|
Returns:
|
|
Résultats du traitement OCR en lot
|
|
"""
|
|
logger.info(f"Traitement OCR en lot de {len(doc_ids)} documents")
|
|
|
|
results = []
|
|
for doc_id in doc_ids:
|
|
try:
|
|
result = process_document_ocr.delay(doc_id, {}).get()
|
|
results.append(result)
|
|
except Exception as e:
|
|
logger.error(f"Erreur lors de l'OCR en lot pour {doc_id}: {e}")
|
|
results.append({
|
|
'doc_id': doc_id,
|
|
'status': 'failed',
|
|
'error': str(e)
|
|
})
|
|
|
|
return {
|
|
'batch_status': 'completed',
|
|
'total_documents': len(doc_ids),
|
|
'results': results
|
|
}
|