// RUN: tf-mlir-translate -mlir-to-graphdef %s -o - | FileCheck %s

func.func @main(%arg0: tensor<*x!tf_type.resource<tensor<8x1xf32>>>) -> tensor<8x1xf32> {
  %0 = tf_executor.graph {
     %outputs, %control = tf_executor.island wraps "tf.ReadVariableOp"(%arg0) : (tensor<*x!tf_type.resource<tensor<8x1xf32>>>) -> tensor<8x1xf32>
     tf_executor.fetch %outputs : tensor<8x1xf32>
  }
  func.return %0 : tensor<8x1xf32>
}

// RUN: tf-mlir-translate -mlir-to-graphdef %s -o - | FileCheck %s

// Verify that attribute T is not set as ref type.
// CHECK-LABEL: name: "Ref_Variable"
// CHECK-LABEL: name: "foo"
// CHECK:      attr {
// CHECK-NEXT:   key: "T"
// CHECK-NEXT:   value {
// CHECK-NEXT:     type: DT_INT32{{$}}
// CHECK-NEXT:   }
// CHECK-NEXT: }

func.func @main() {
  tf_executor.graph {

// RUN: tf-mlir-translate -mlir-tf-to-hlo-text %s -tf-input-shapes=128,10:10,1024:128,1024 -tf-xla-emit-use-tuple-args -tf-xla-emit-return-tuple | FileCheck %s

module attributes {tf.versions = {bad_consumers = [], min_consumer = 0 : i32, producer = 351 : i32}} {

// RUN: tf-mlir-translate -mlir-to-graphdef %s -o - | FileCheck %s

func.func @main(%arg0: tensor<10xi32>, %arg1: tensor<10xi32>) -> tensor<10xi32>
attributes {tf.entry_function = {inputs = "foo,bar", outputs = "Add"}} {
  %graph = tf_executor.graph {
    // This node would be renamed to bar1 [note: if imported from TF graphdef this would not be possible]

# RUN: tf-mlir-translate -graphdef-to-mlir -tf-enable-shape-inference-on-import=false %s -tf-graph-as-function -o - | FileCheck %s

# Verify arg and ret devices are added as arg and ret attributes.

# CHECK-LABEL: func @main
# CHECK-SAME:  (%[[ARG_0:[a-z0-9]+]]: tensor<*xf32> {tf.device = "/CPU:0"}, %[[ARG_1:[a-z0-9]+]]: tensor<2x4x6x8xi32>) -> (tensor<*xf32>, tensor<*xi32> {tf.device = "/CPU:1"})

node {
  name: "args_0"
  op: "_Arg"
  device: "/CPU:0"

// RUN: tf-tfrt-opt -pass-pipeline='builtin.module(tf-executor-to-tfrt-pipeline{target-tpurt=true})' %s | FileCheck %s

module attributes {tf_saved_model.semantics}  {
  // CHECK-LABEL: func @main
  func.func @main_func() -> (tensor<*xf32> {tf_saved_model.index_path = ["a"]}) attributes {tf_saved_model.exported_names = ["main_func"]} {
    %0 = tf_executor.graph {

// RUN: tf-tfrt-opt -tf-restore-pruning %s | FileCheck %s

// CHECK-LABEL:   func.func @prune_unused_restore
func.func @prune_unused_restore() {
  %cst = "tf.Const"() <{value = dense<"restore_ariables"> : tensor<!tf_type.string>}> : () -> tensor<!tf_type.string>
  %cst_0 = "tf.Const"() <{value = dense<""> : tensor<1x!tf_type.string>}> : () -> tensor<1x!tf_type.string>

// RUN: flatbuffer_translate -mlir-to-tflite-flatbuffer %s -o - | flatbuffer_to_string - | FileCheck %s

// CHECK: {
// CHECK-NEXT:   version: 3,
// CHECK-NEXT:  operator_codes: [  ],
// CHECK-NEXT:  subgraphs: [ {
// CHECK-NEXT:    tensors: [ {
// CHECK-NEXT:      shape: [  ],
// CHECK-NEXT:      type: VARIANT,
// CHECK-NEXT:      buffer: 1,
// CHECK-NEXT:      name: "arg0",
// CHECK-NEXT:      quantization: {
// CHECK-EMPTY:

// RUN: odml-to-stablehlo-opt %s -unfold-splat-constant-pass -cse -verify-diagnostics | FileCheck %s

// CHECK-LABEL: @unfold_splat_constant_float
func.func @unfold_splat_constant_float() -> tensor<1x750xf32> {
  %cst = mhlo.constant dense<7.680000e+02> : tensor<1x750xf32>
  func.return %cst : tensor<1x750xf32>

  // CHECK-DAG: %0 = mhlo.constant dense<7.680000e+02> : tensor<f32>

// RUN: tf-mlir-translate -mlir-to-graphdef %s -o - | FileCheck %s

func.func @main() {
^bb0:
  // CHECK:      key: "listvalue"
  // CHECK-NEXT: value {
  // CHECK-NEXT:   list {
  // CHECK-NEXT:     s: " \n"
  // CHECK-NEXT:   }
  // CHECK:      key: "value"
  // CHECK-NEXT: value {
  // CHECK-NEXT:   s: " 0\n\000\000"
  tf_executor.graph {