Create new canned policy file `getonly.json`. This policy enables users to download all objects under `my-bucketname`.

```json
cat > getonly.json << EOF
{
  "Version": "2012-10-17",
  "Statement": [
	{
	  "Action": [
		"s3:GetObject"
	  ],
	  "Effect": "Allow",
	  "Resource": [

E também teremos anotações/documentação em conformidade.

## Modelos aninhados

Cada atributo de um modelo Pydantic tem um tipo.

Mas esse tipo pode ser outro modelo Pydantic.

Portanto, você pode declarar "objects" JSON profundamente aninhados com nomes, tipos e validações de atributos específicos.

Tudo isso, aninhado arbitrariamente.

### Defina um sub-modelo

Por exemplo, nós podemos definir um modelo `Image`:

So, if we create a Pydantic object `user_in` like:

```Python
user_in = UserIn(username="john", password="secret", email="******@****.***")
```

and then we call:

```Python
user_dict = user_in.dict()
```

we now have a `dict` with the data in the variable `user_dict` (it's a `dict` instead of a Pydantic model object).

And if we call:

```Python
print(user_dict)
```

// since most usage will be by language specific wrappers.
//
// Conventions:
// * We use the prefix TF_ for everything in the API.
// * Objects are always passed around as pointers to opaque structs
//   and these structs are allocated/deallocated via the API.
// * TF_Status holds error information.  It is an object type
//   and therefore is passed around as a pointer to an opaque
//   struct as mentioned above.

        if (typeof options.startDate === 'object')
            this.startDate = moment(options.startDate);

        if (typeof options.endDate === 'object')
            this.endDate = moment(options.endDate);

        if (typeof options.minDate === 'object')
            this.minDate = moment(options.minDate);

        if (typeof options.maxDate === 'object')
            this.maxDate = moment(options.maxDate);

```

In diesem Beispiel haben wir kein Pydantic-Modell deklariert. Tatsächlich wird der Requestbody nicht einmal als JSON <abbr title="von einem einfachen Format, wie Bytes, in Python-Objekte konvertieren">geparst</abbr>, sondern direkt als `bytes` gelesen und die Funktion `magic_data_reader ()` wäre dafür verantwortlich, ihn in irgendeiner Weise zu parsen.

	if !o.isSorted() {
		t.Fatal("Expected sorted objects")
	}

	// Swap first and last
	o[0], o[len(o)-1] = o[len(o)-1], o[0]
	if o.isSorted() {
		t.Fatal("Expected unsorted objects")
	}

	sorted := o.sort()
	if !o.isSorted() {
		t.Fatal("Expected sorted o objects")
	}
	if !sorted.entries().isSorted() {
		t.Fatal("Expected sorted wrapped objects")
	}
	want := loadMetacacheSampleNames

        "it.unimi.dsi.fastutil.objects.Object2IntMap",
        "it.unimi.dsi.fastutil.objects.Object2IntOpenHashMap",
        "it.unimi.dsi.fastutil.objects.Object2ObjectLinkedOpenHashMap",
        "it.unimi.dsi.fastutil.objects.Object2ObjectOpenCustomHashMap",
        "it.unimi.dsi.fastutil.objects.Object2ObjectOpenHashMap",
        "it.unimi.dsi.fastutil.objects.ObjectOpenCustomHashSet",

TF_CAPI_EXPORT extern void TFE_ContextEndStep(TFE_Context* ctx);

#ifdef __cplusplus
} /* end extern "C" */
#endif

#ifdef __cplusplus
// A workaround to ease conversion to and from numpy objects and
// TFE_TensorHandle's.
//
// TODO(ashankar): Figure out an alternative scheme that precludes the need for
// these API-boundary breaking methods.
namespace tensorflow {
class Tensor;

    @Inject
    public GradleDocumentationExtension(ObjectFactory objects) {
        releaseNotes = objects.newInstance(ReleaseNotes.class);
        userManual = objects.newInstance(UserManual.class);
        dslReference = objects.newInstance(DslReference.class);
        javadocs = objects.newInstance(Javadocs.class);
        kotlinDslReference = objects.newInstance(KotlinDslReference.class);
    }

    /**