}
    std::vector<TensorHandlePtr> components;
    components.reserve(inputs.size());
    for (int i = 0; i < inputs.size(); ++i) {
      if (absl::holds_alternative<ParallelTensor*>(inputs[i])) {
        std::string message(absl::StrCat(
            "Expected all inputs to TPUReplicatedInput to be non-parallel "
            "TensorHandles. The input ",
            i,
            " was a parallel tensor (already "

//   return tf.exp(input), grad
// outputs = [f(inputs[0])]
Status ExpWithPassThroughGrad(AbstractContext* ctx,
                              absl::Span<AbstractTensorHandle* const> inputs,
                              absl::Span<AbstractTensorHandle*> outputs) {
  Tape tape(/*persistent=*/false);
  tape.Watch(inputs[0]);  // Watch x.
  AbstractTensorHandle* exp_output;
  TF_RETURN_IF_ERROR(ops::Exp(ctx, inputs[0], &exp_output, "Exp"));

	"istio.io/istio/pkg/workloadapi/security"
)

func TestPodWorkloads(t *testing.T) {
	cases := []struct {
		name   string
		inputs []any
		pod    *v1.Pod
		result *workloadapi.Workload
	}{
		{
			name:   "simple pod not running and not have podIP",
			inputs: []any{},
			pod: &v1.Pod{
				TypeMeta: metav1.TypeMeta{},
				ObjectMeta: metav1.ObjectMeta{
					Name:      "name",
					Namespace: "ns",

        "inputs.file('a')"                                          | "TaskInputs.file(Object)"
        "inputs.files('a')"                                         | "TaskInputs.files(Object...)"
        "inputs.dir('a')"                                           | "TaskInputs.dir(Object)"
        "inputs.property('key', 'value')"                           | "TaskInputs.property(String, Object)"

func (f *Selector) Selects(inputs Set) bool {
	return !inputs.containsAny(f.absent) && inputs.containsAll(f.present)
}

// Excludes returns false, if the given set of labels, even combined with new ones, could end up satisfying the Selector.
// It returns false, if Matches would never return true, even if new labels are added to the input set.
func (f *Selector) Excludes(inputs Set) bool {
	return inputs.containsAny(f.absent)
}

     *
     * @since 6.0
     */
    @TaskAction
    protected void compile(InputChanges inputs) {
        DefaultJavaCompileSpec spec = createSpec();
        if (!compileOptions.isIncremental()) {
            performFullCompilation(spec);
        } else {
            performIncrementalCompilation(inputs, spec);
        }
    }

  }
  func.return
}

// CHECK:      [[INPUT_A:%.+]], {{%.+}} = tf_executor.island {
// CHECK-NEXT:   [[OP_A_OUTPUT:%.+]] = "tf.opA"([[ARG_0]])
// CHECK-NEXT:   tf_executor.yield [[OP_A_OUTPUT]] :
// CHECK:      [[ISLAND_0_OUTPUT:%.+]], {{%.+}} = tf_executor.island {
// CHECK-NEXT:   [[OP_B_OUTPUT:%.+]] = "tf.opB"([[INPUT_A]]) {_parallel_execution_ids = "p0:0"}
// CHECK:        tf_executor.yield [[OP_B_OUTPUT]] :

    // Fetch inputs from the OpKernelContext. Inputs are the same as the ones
    // for XlaCompile, except that the must-be-constant inputs that appear in
    // the beginning are stripped off and the closure key is appended as the
    // last input. So the inputs look like: input tensors, resource variables,
    // closure key tensor.
    std::vector<const Tensor*> inputs = InputsFromContext(ctx);

        file("symlink2").createLink(actualDir)

        buildFile << """
            task myTask {
                def outputFile = file("build/output.txt")
                inputs.dir("symlink1")
                inputs.dir("symlink2")
                outputs.file(outputFile)

                doLast {
                    outputFile.text = "Hello world"
                }
            }
        """

  // The first one will the function's output
  std::vector<TF_Output> outputs;

  // Add while loop to func_graph_
  {
    // The inputs to the while loop
    std::vector<TF_Output> inputs = {{feed1, 0}, {feed2, 0}};
    std::unique_ptr<TF_WhileParams> params(new TF_WhileParams(
        TF_NewWhile(func_graph_, &inputs[0], inputs.size(), s_)));