n8n Automation Integration

@shared_task(name='documentos.tasks.extract_document_metadata')
def extract_document_metadata(revision_id):
    """
    Tarea para extraer texto y metadatos de un documento (PDF, etc.)
    """
    from .models import Revision
    from django.conf import settings
    import requests
    
    try:
        revision = Revision.objects.get(pk=revision_id)
        revision.estado_extraccion = 'PROCESANDO'
        revision.save()

        # Prepare callback URL for n8n to send results back
        base_callback_url = settings.INTERNAL_SITE_URL
        internal_callback_url = f"{base_callback_url}/documentos/api/callback-procesamiento/{revision.id}/"

        payload = {
            'revision_id': revision.id,
            'documento_id': revision.documento.id,
            'filename': os.path.basename(revision.archivo.name),
            'file_url': revision.archivo.url,  # MinIO signed URL
            'file_key': revision.archivo.name,
            'tipo_documento': revision.documento.tipo_documento.nombre,
            'callback_url': internal_callback_url,
            'metadatos_requeridos': list(
                revision.documento.tipo_documento.metadatos_config.values_list('nombre', flat=True)
            )
        }
        
        n8n_url = settings.N8N_PROCESS_DOCUMENT_WEBHOOK_URL
        response = requests.post(n8n_url, json=payload, timeout=10)
        logger.info(f"Revision {revision_id}: Enviada a n8n. Status: {response.status_code}")
        
        return True

    except Exception as e:
        logger.error(f"Error en extract_document_metadata: {str(e)}")
        return False

{
  "name": "Process Document - PDF + Metadata",
  "nodes": [
    {
      "name": "Webhook",
      "type": "n8n-nodes-base.webhook",
      "parameters": {
        "path": "process-document",
        "responseMode": "onReceived"
      }
    },
    {
      "name": "Download PDF",
      "type": "n8n-nodes-base.httpRequest",
      "parameters": {
        "url": "={{$json.file_url}}",
        "method": "GET",
        "responseFormat": "file"
      }
    },
    {
      "name": "Extract Text (PyMuPDF)",
      "type": "n8n-nodes-base.code",
      "parameters": {
        "language": "python",
        "code": "import fitz\n\npdf_data = items[0]['binary']['data']\ndoc = fitz.open(stream=pdf_data)\n\ntext = ''\nfor page in doc:\n    text += page.get_text()\n\nreturn {'texto': text, 'num_pages': len(doc)}"
      }
    },
    {
      "name": "Extract Metadata (Gemini)",
      "type": "n8n-nodes-base.httpRequest",
      "parameters": {
        "url": "https://generativelanguage.googleapis.com/v1beta/models/gemini-pro:generateContent",
        "method": "POST",
        "authentication": "genericCredentialType",
        "headers": {
          "Content-Type": "application/json"
        },
        "body": {
          "contents": [
            {
              "parts": [
                {
                  "text": "Extract these fields from the document: {{$json.metadatos_requeridos}}. Document text: {{$json.texto}}"
                }
              ]
            }
          ]
        }
      }
    },
    {
      "name": "Callback to Django",
      "type": "n8n-nodes-base.httpRequest",
      "parameters": {
        "url": "={{$json.callback_url}}",
        "method": "POST",
        "body": {
          "texto": "={{$json.texto}}",
          "metadata": "={{$json.gemini_response}}",
          "num_paginas": "={{$json.num_pages}}"
        }
      }
    }
  ]
}

from django.views.decorators.csrf import csrf_exempt
from django.http import JsonResponse
import json

@csrf_exempt
def callback_procesamiento(request, revision_id):
    """
    Callback endpoint para recibir resultados de n8n
    """
    if request.method == 'POST':
        try:
            data = json.loads(request.body)
            revision = Revision.objects.get(pk=revision_id)
            
            # Update extracted text
            if 'texto' in data:
                revision.documento.contenido_texto = data['texto']
                revision.documento.save()
            
            # Update metadata
            if 'metadata' in data:
                revision.datos_extraidos = data['metadata']
            
            revision.estado_extraccion = 'COMPLETADO'
            revision.save()
            
            return JsonResponse({'status': 'success'})
        except Exception as e:
            logger.error(f"Error en callback: {e}")
            return JsonResponse({'error': str(e)}, status=500)
    
    return JsonResponse({'error': 'Method not allowed'}, status=405)