fix: propagate dynamic IPs/ports to service pages; add apply restart feedback

Service pages (Email, Calendar, Files) now read IPs and ports from the
config API instead of hardcoded 172.20.0.x constants:
- GET /api/config now includes service_ips (dns, vip_mail, vip_calendar,
  vip_files, vip_webdav) computed from ip_range via ip_utils
- Email.jsx: mailIp, dnsIp, imapPort, smtpPort, webmailPort from context
- Calendar.jsx: calendarIp, dnsIp, calendarPort from context
- Files.jsx: filesIp, webdavIp, webdavPort, filegatorPort from context

Apply button now shows restart progress:
- "Restarting containers — please wait…" spinner while polling /health
- "Containers restarted successfully" on success (clears after 4s)
- "Timed out" / error message if health doesn't come back in 45s

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

api/app.py

@@ -400,6 +400,16 @@ def get_config():
             'ip_range': identity.get('ip_range', os.environ.get('CELL_IP_RANGE', '172.20.0.0/16')),
             'wireguard_port': identity.get('wireguard_port', int(os.environ.get('WG_PORT', '51820'))),
         }
+        # Expose computed per-service IPs so the frontend doesn't need to derive them
+        import ip_utils as _ip_utils_cfg
+        _ips = _ip_utils_cfg.get_service_ips(config['ip_range'])
+        config['service_ips'] = {
+            'dns':          _ips['dns'],
+            'vip_mail':     _ips['vip_mail'],
+            'vip_calendar': _ips['vip_calendar'],
+            'vip_files':    _ips['vip_files'],
+            'vip_webdav':   _ips['vip_webdav'],
+        }
         config['service_configs'] = service_configs
         return jsonify(config)
     except Exception as e: