FASE 4-5-6: Infraestructura, CRM, Service Orders, Notificaciones, Ahorro, Logistica, API Publica

FASE 4:
- Redis cache de stock con fallback graceful
- Multi-moneda (MXN/USD) con contabilidad en MXN
- Proveedores y ordenes de compra completo
- Meilisearch 1.5M+ partes indexadas
- Metabase KPIs con dashboard auto-generado

FASE 5:
- CRM mejorado: activities, tags, loyalty program, analytics
- Imagenes de partes: upload, resize, thumbnails WebP
- Ordenes de servicio Kanban: received->diagnosis->repair->ready->delivered
- Garantias/RMA, alertas de reorden, multi-sucursal
- Stubs BNPL (APLAZO) y ERP Sync (Aspel/Contpaqi)

FASE 6:
- Notificaciones automaticas: push/WhatsApp/email/in-app
- Reportes de ahorro vs retail_price
- Logistica + tracking: DHL, FedEx, Estafeta, 99min, Uber
- API Publica: API keys, rate limiting, catalog search

Migraciones: v1.9-v3.0
Tests: 93/93 pasando
Backup: nexus_backup_20260427_045859.tar.gz

scripts/sync_meilisearch.py

@@ -0,0 +1,92 @@
+#!/usr/bin/env python3
+"""Bulk-sync parts from PostgreSQL master DB into Meilisearch.
+
+Usage:
+    python3 scripts/sync_meilisearch.py [--clear]
+
+Requires environment variables:
+    MASTER_DB_URL=postgresql://user:pass@localhost/nexus_autoparts
+"""
+
+import os
+import sys
+import argparse
+
+sys.path.insert(0, os.path.join(os.path.dirname(os.path.dirname(os.path.abspath(__file__))), 'pos'))
+
+import psycopg2
+from services.meili_search import ensure_index, index_parts_bulk, clear_index, health_check
+
+
+def fetch_parts(conn, batch_size=5000):
+    """Yield parts from PostgreSQL as dicts."""
+    cur = conn.cursor(name='parts_cursor')
+    cur.execute("""
+        SELECT id_part, oem_part_number, name_part, name_es,
+               description, description_es, image_url, group_id
+        FROM parts
+        ORDER BY id_part
+    """)
+    while True:
+        rows = cur.fetchmany(batch_size)
+        if not rows:
+            break
+        for row in rows:
+            yield {
+                'id_part': row[0],
+                'oem_part_number': row[1],
+                'name_part': row[2],
+                'name_es': row[3] or row[2],
+                'description': row[4] or '',
+                'description_es': row[5] or '',
+                'image_url': row[6] or '',
+                'group_id': row[7],
+            }
+    cur.close()
+
+
+def main():
+    parser = argparse.ArgumentParser(description='Sync parts to Meilisearch')
+    parser.add_argument('--clear', action='store_true', help='Clear index before sync')
+    parser.add_argument('--batch-size', type=int, default=5000, help='PostgreSQL fetch batch size')
+    parser.add_argument('--index-batch', type=int, default=1000, help='Meilisearch upload batch size')
+    args = parser.parse_args()
+
+    print("Meilisearch Sync")
+    print("=" * 50)
+
+    if not health_check():
+        print("ERROR: Meilisearch is not reachable.")
+        print(f"  URL: {os.environ.get('MEILI_URL', 'http://localhost:7700')}")
+        sys.exit(1)
+
+    master_db_url = os.environ.get('MASTER_DB_URL')
+    if not master_db_url:
+        print("ERROR: MASTER_DB_URL environment variable is required.")
+        sys.exit(1)
+
+    ensure_index()
+
+    if args.clear:
+        print("Clearing existing index...")
+        clear_index()
+
+    print(f"Connecting to PostgreSQL...")
+    conn = psycopg2.connect(master_db_url)
+
+    # Count total
+    cur = conn.cursor()
+    cur.execute("SELECT COUNT(*) FROM parts")
+    total_rows = cur.fetchone()[0]
+    cur.close()
+    print(f"Parts to index: {total_rows}")
+
+    print("Indexing...")
+    indexed = index_parts_bulk(fetch_parts(conn, args.batch_size), batch_size=args.index_batch)
+    conn.close()
+
+    print(f"Done. Indexed {indexed} documents.")
+
+
+if __name__ == '__main__':
+    main()