}

func main() {
	testbasic()
	testfloat()
	testnan()
}

func testbasic() {
	// Test a map literal.
	mlit := map[string]int{"0": 0, "1": 1, "2": 2, "3": 3, "4": 4}
	for i := 0; i < len(mlit); i++ {
		s := string([]byte{byte(i) + '0'})
		if mlit[s] != i {
			panic(fmt.Sprintf("mlit[%s] = %d\n", s, mlit[s]))
		}
	}

	mib := make(map[int]bool)
	mii := make(map[int]int)
	mfi := make(map[float32]int)

  RegisterPasses();
  mlir::DialectRegistry registry;
  mlir::RegisterAllTensorFlowDialects(registry);
  mlir::stablehlo::registerAllDialects(registry);
  registry.insert<mlir::shape::ShapeDialect>();
  mlir::MLIRContext context(registry);
  mlir::OwningOpRef<mlir::ModuleOp> module;
  {
    mlir::StatusScopedDiagnosticHandler diagnostic_handler(&context);

// CHECK-LABEL: @main
// CHECK-SAME: ([[f0:%.*]]: !mlrt.async_handle, [[f1:%.*]]: !mlrt.async_handle, [[f2:%.*]]: !mlrt.async_handle)
func.func @main(%f0: !mlrt.async_handle, %f1: !mlrt.async_handle, %f2: !mlrt.async_handle) -> () {
  // CHECK-NEXT: mlrt.await_all_handle [[f0]], [[f1]], [[f2]]
  // CHECK-NOT: mlrt.await_handle
  // CHECK-NEXT: return
  mlrt.await_handle %f0
  mlrt.await_handle %f1
  mlrt.await_handle %f2
  func.return
}

    (!mlrt.promise, !mlrt.promise, !mlrt.promise,
     !mlrt.future, !mlrt.future, !mlrt.future)

  // CHECK: [[handle_0:%.*]] = mlrt.async([[input0]], [[promises]]#0)
  // CHECK-SAME: callee = @main_stream_0
  %handle_0 = mlrt.async(%input0, %promise_b)
    {callee = @main_stream_0} :
    (tensor<i32>, !mlrt.promise) -> !mlrt.async_handle

  %1:9 = "tf.While"(%cst_0, %max_iterations, %array_handle, %array_flow, %matrix , %array_handle, %array_flow, %matrix, %bound) {body= @"sort_map/while_body", cond = @"sort_map/while_cond",...

// CHECK: mlrt.async([[FUTURES]]#0) {callee = @main_stream_1
// CHECK: mlrt.await_handle
// CHECK: mlrt.await_handle
// CHECK: mlrt.await_handle
// CHECK: mlrt.await_handle

func.func @main() {
  "tf.DummySideEffecting"() {id = 1} : () -> ()
  "tf.DummySideEffecting"() {id = 2} : () -> ()
  "tf.DummySideEffecting"() {id = 3} : () -> ()

// Migrated from previous SWIG file (mlir.i) authored by aminim@.
#ifndef TENSORFLOW_COMPILER_MLIR_PYTHON_MLIR_H_
#define TENSORFLOW_COMPILER_MLIR_PYTHON_MLIR_H_

#include <string>
#include <vector>

#include "absl/strings/string_view.h"
#include "tensorflow/c/eager/c_api.h"
#include "tensorflow/c/tf_status.h"

namespace tensorflow {

// Simple wrapper to support tf.mlir.experimental.convert_graph_def.

// RUN: tf-mlir-translate -mlir-tf-to-hlo-text %s -tf-input-shapes=: -tf-xla-emit-return-tuple | FileCheck %s
// RUN: tf-mlir-translate -mlir-tf-to-hlo-text %s -tf-input-shapes=: -tf-xla-emit-use-tuple-args -tf-xla-emit-return-tuple | FileCheck -check-prefix=TUPLE-ARGS %s
// RUN: tf-mlir-translate -mlir-tf-to-hlo-text %s -tf-input-shapes=: | FileCheck -check-prefix=NO_RET_TUPLE %s

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

// Verifies that optional symbol ref attributes that aren't optional in TensorFlow are handled by setting the value to an empty string.
module attributes {tf.versions = {bad_consumers = [], min_consumer = 12 : i32, producer = 458 : i32}} {
  func.func @main() {
    tf_executor.graph {

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

// Test graphdef export is producing valid GraphDef. The test imports the
// generated GraphDef to make sure the export is valid.

module attributes {tf.versions = {bad_consumers = [], min_consumer = 12 : i32, producer = 890 : i32}}  {