// Test simple functions.

package main

func assertequal(is, shouldbe int, msg string) {
	if is != shouldbe {
		print("assertion fail", msg, "\n")
		panic(1)
	}
}

func f1() {
}

func f2(a int) {
}

func f3(a, b int) int {
	return a + b
}

func f4(a, b int, c float32) int {
	return (a+b)/2 + int(c)
}

func f5(a int) int {
	return 5
}

func f6(a int) (r int) {

# RUN: tf-mlir-translate -graphdef-to-mlir %s

node {
  name: "input"
  op: "Placeholder"
  attr {
    key: "dtype"
    value {
      type: DT_BOOL
    }
  }
}

node {
  name: "func0"
  op: "func_name"
  input: "input"
}

library {
  function {
    signature {
      name: "func_name"
      input_arg {
        name: "arg0"
        type: DT_BOOL
      }
    }
    ret {

#include <string>
#include <utility>

#include "llvm/ADT/StringExtras.h"
#include "mlir/Dialect/Func/IR/FuncOps.h"  // from @llvm-project

namespace mlir {
namespace quant {

constexpr StringRef kTfEntryFunctionAttr = "tf.entry_function";

void AddEntryFunctionInput(StringRef input_name, func::FuncOp func_op) {
  auto entry_func_attr =
      func_op->getAttrOfType<DictionaryAttr>(kTfEntryFunctionAttr);
  if (!entry_func_attr) return;

// CHECK-LABEL: func @move_broadcast
func.func @move_broadcast(%arg0: tensor<f32>) -> () {
  // CHECK:      %[[ELEM_0:.*]] = "tf.Const"()
  // CHECK:      {ici_weight_distribution_mlir_bridge_marker = true}
  // CHECK-NEXT: %[[SHAPE_0:.*]] = "tf.Const"()
  // CHECK:      {ici_weight_distribution_mlir_bridge_marker = true}

  ~TFConcreteFunction() override = default;

 private:
  TFConcreteFunction(std::unique_ptr<FlatTensorFunction> func,
                     FunctionMetadata metadata);

  TFConcreteFunction(const TFConcreteFunction&) = delete;
  TFConcreteFunction& operator=(const TFConcreteFunction&) = delete;

  std::unique_ptr<FlatTensorFunction> func_;
  FunctionMetadata metadata_;
};

}  // namespace tensorflow

};

// Folds tf.IdentityOp and tf.IdentityNOp if op device and the argument devices
// from the defining ops match.
void TensorDeviceCopyConversionPass::runOnOperation() {
  func::FuncOp func_op = getOperation();

  auto should_fold_op_func = [&func_op](const Value &arg,
                                        const StringAttr &op_device) {
    // In TFRT TPU, tensor transfer is handled specifically by D2H and

#define DEBUG_TYPE "quantization-context"

namespace mlir {
namespace quant {

QuantizeContext::QuantizeContext(func::FuncOp func, const DeviceTarget &spec)
    : func_(func), target_spec_(spec) {
  llvm::DenseMap<Value, int> value_to_state;
  func.walk([&](quantfork::QuantizeRegionOp op) {
    for (int i = 0, e = op.getNumOperands(); i != e; ++i) {

    // Builder builder(func_op.getContext());
    func_op->setAttr(
        kTfInputShapesAttr,
        ArrayAttr::get(func_op.getContext(), updated_input_shapes_attr));
  }
}

// Validates func ops. Returns `failure` if the function is invalid.
LogicalResult ValidateFuncOp(func::FuncOp func_op) {
  auto input_shapes_attr =
      func_op->getAttrOfType<ArrayAttr>(kTfInputShapesAttr);

  const StringSet<> noinline_functions = GetNoinlineFunctionsSet();

  func::FuncOp func_op = getOperation();
  Builder builder(&getContext());

  // Adds the `tf._noinline = true` attribute to the function if the name
  // matches.
  if (noinline_functions.contains(func_op.getSymName())) {
    func_op->setAttr(kTfNoinlineAttr, builder.getBoolAttr(true));
    LLVM_DEBUG(llvm::dbgs()

### Passo 4: defina uma **função de rota**

Esta é a nossa "**função de rota**":

* **rota**: é `/`.
* **operação**: é `get`.
* **função**: é a função abaixo do "decorador" (abaixo do `@app.get("/")`).

```Python hl_lines="7"
{!../../../docs_src/first_steps/tutorial001.py!}
```

Esta é uma função Python.