TF_Operation** return_opers;
  int num_return_opers;
  TF_ImportGraphDefResultsReturnOperations(results, &num_return_opers,
                                           &return_opers);
  ASSERT_EQ(1, num_return_opers);
  EXPECT_EQ(scalar2, return_opers[0]);  // not remapped

  TF_DeleteImportGraphDefResults(results);

  // Import again, with control dependencies, into the same graph.
  TF_DeleteImportGraphDefOptions(opts);

{* ../../docs_src/body/tutorial001_py310.py hl[16] *}

...и укажите тип параметра как созданную вами модель, `Item`.

## Результаты { #results }

Всего лишь с этой аннотацией типов Python **FastAPI**:

* Считает тело запроса как JSON.
* Приведёт данные к соответствующим типам (если потребуется).
* Проведёт валидацию данных.

		DiskID:   *client.diskID.Load(),
		ScanMode: int(scanMode),
		Cache:    &cache,
	})
	if err != nil {
		return cache, toStorageErr(err)
	}
	var final *dataUsageCache
	err = st.Results(func(resp *nsScannerResp) error {
		if resp.Update != nil {
			select {
			case <-ctx.Done():
			case updates <- *resp.Update:
			}
		}
		if resp.Final != nil {
			final = resp.Final
		}

Vous écrivez le premier paramètre, `first_name`, puis un point (`.`) et appuyez sur `Ctrl+Espace` pour déclencher l'auto-complétion.

Mais malheureusement, rien d'utile n'en résulte :

<img src="/img/python-types/image01.png">

### Ajouter des types

Modifions une seule ligne de la version précédente.

Nous allons changer seulement cet extrait, les paramètres de la fonction, de :

{* ../../docs_src/body/tutorial001_py310.py hl[16] *}

...y declara su tipo como el modelo que creaste, `Item`.

## Resultados { #results }

Con solo esa declaración de tipo en Python, **FastAPI** hará lo siguiente:

* Leer el body del request como JSON.
* Convertir los tipos correspondientes (si es necesario).
* Validar los datos.

			continue
		}
		diskCount++
	}
	return diskCount
}

// hashOrder - hashes input key to return consistent
// hashed integer slice. Returned integer order is salted
// with an input key. This results in consistent order.
// NOTE: collisions are fine, we are not looking for uniqueness
// in the slices returned.
func hashOrder(key string, cardinality int) []int {
	if cardinality <= 0 {

## Overview

Replication relies on immutability provided by versioning to sync objects between the configured source and replication target. Replication results in the object data, metadata, last modification time and version ID all being identical between the source and target. Thus version ordering is automatically guaranteed on the source and target clusters.

		}

		for p := range latestMeta.Parts {
			if verifyErr != nil {
				dataErrsByPart[p][i] = convPartErrToInt(verifyErr)
			} else {
				dataErrsByPart[p][i] = verifyResp.Results[p]
			}
		}
	}

	// Build dataErrs by disk from dataErrs by part
	for part, disks := range dataErrsByPart {
		for disk := range disks {
			dataErrsByDisk[disk][part] = dataErrsByPart[part][disk]
		}

### Read Heroes { #read-heroes }

We can **read** `Hero`s from the database using a `select()`. We can include a `limit` and `offset` to paginate the results.

{* ../../docs_src/sql_databases/tutorial001_an_py310.py ln[48:55] hl[51:52,54] *}

### Read One Hero { #read-one-hero }

We can **read** a single `Hero`.

//     op for shape inference.
//   - The types need not be set in `input_shapes` as it is not used.
//   - The number of `input_tensors` should be the same as the number of items
//     in `input_shapes`.
//
// The results are returned in `output_shapes` and
// `output_resource_shapes_and_types`. The caller is responsible for freeing the
// memory in these buffers by calling `TF_DeleteShapeAndTypeList`.
TF_CAPI_EXPORT extern void TFE_InferShapes(