jcifs-ng will be a proper choice for many users. 
There are a lot of SMB devices in the world.
Some of them only work with the old jcifs library.
If you want to support many SMB devices, CodeLibs jcifs library will be helpful.
For example, since [Fess](https://github.com/codelibs/fess) needs to support many SMB devices, it uses this library.

   *
   * @since 8.0
   */
  public static final byte DLE = 16;

  /**
   * Device Control 1. Characters for the control of ancillary devices associated with data
   * processing or telecommunication systems, more especially switching devices "on" or "off." (If a
   * single "stop" control is required to interrupt or turn off ancillary devices, DC4 is the
   * preferred assignment.)
   *
   * @since 8.0
   */

                {n = 2 : i32,
                 devices = {DEVICE_ALIAS_0 = ["/DEVICE:0", "/DEVICE:1"],
                            DEVICE_ALIAS_1 = ["/DEVICE:2", "/DEVICE:3"]}} {
  // Inside the region, %0, %2, %4, and %6 corresponds to
  // "/DEVICE:0"/"/DEVICE:2" and %1, %3, %5, and %7 corresponds to
  // "/DEVICE:1"/"/DEVICE:3", depending on which device alias is used.
  %k = "tf_device.launch"() ( {

// RUN: tf-opt -split-input-file -verify-diagnostics %s -tf-replica-id-to-device-ordinal | FileCheck %s


// Tests device ordinal is set correctly for multiple devices.
// CHECK-LABEL: func @device_ordinal_attr_added_multiple_devices
module attributes {tf.devices = ["/job:worker/replica:0/task:0/device:CPU:0", "/job:worker/replica:0/task:0/device:TPU_SYSTEM:0", "/job:worker/replica:0/task:0/device:TPU:0", "/job:worker/replica:0/task:0/device:TPU:1"]} {

       tf_device.return
    }) {device = "TPU_REPLICATED_CORE_0"} : () -> ()
    "tf_device.launch"() ({
      // CHECK:  "tf.B"(%[[RESOURCE_OUT]])
      "tf.B"(%1) : (tensor<4xf32>) -> ()
       tf_device.return
    }) {device = "TPU_REPLICATED_CORE_0"} : () -> ()
    tf_device.return
  }
  func.return
}

// Tests AssignVariable op using composite device resource is wrapped inside
// tf_device.Cluster.

  }

  // Call AddDevices to register the XLA devices.
  //
  // It may be worth refactoring out XlaOpRegistry::RegisterCompilationDevice to
  // make this more direct, but probably not worth it solely for this test.
  std::vector<std::unique_ptr<Device>> devices;
  TF_RETURN_IF_ERROR(DeviceFactory::AddDevices(session_options, "", &devices));

  GraphOptimizationPassOptions opt_options;
  opt_options.graph = graph;

							"vendor.com/device=device1",
							"vendor.com/device=device2",
						},
						"test-plugin2": {
							"vendor.com/device=device1",
							"vendor.com/device=device2",
						},
					},
				},
			},
			expectedResult: []kubecontainer.CDIDevice{
				{
					Name: "vendor.com/device=device1",
				},
				{
					Name: "vendor.com/device=device1",
				},
				{

// configuring the persistor used in the DeviceCompiler. Please note that
// non-XLA devices aren't supported yet. This is because:
// 1. PjRtClient doesn't support data transfer for non-XLA devices yet
// 2. Fetching the PjRtClient for non-XLA devices is also not supported yet
Status GetOrCreatePjRtDeviceCompilerAndProfiler(
    const OpKernelContext& ctx, const XlaPlatformInfo& platform_info,

// such as TPUExecute or XlaExecute depending on the device type and specific
// host runtime. Also does some optimization. Will return an error if it fails.
// The output Runtime ops depends on both Device Type and Runtime Host.
//
// Input:
//     Tensorflow Dialect MLIR with tf_device.cluster ops and virtual devices.
//     xla_device_type - The device type that is being targeted.
// Output:

// RUN: tf-opt %s -split-input-file -tf-xla-rewrite | FileCheck %s


module attributes {tf.devices = ["/job:worker/replica:0/task:0/device:CPU:0", "/job:worker/replica:0/task:0/device:GPU:0"]} {
  // CHECK-LABEL: func.func @convert_cluster_func
  func.func @convert_cluster_func(%arg0: tensor<i32>) -> tensor<i32> {
    // CHECK: "tf.XlaLaunch"(%arg0) <{function = @func, operandSegmentSizes = array<i32: 0, 1, 0>}> : (tensor<i32>) -> tensor<i32>