return err
	}

	// For *any* failures after calling `CreateInpodRules`, we must return PartialAdd error.
	// The pod was injected with iptables rules, so it must be annotated as "inpod" - even if
	// the following fails.
	// This is so that if it is removed from the mesh, the inpod rules will unconditionally
	// be removed.

	log.Debug("notifying subscribed node proxies")

And whenever you output that data, even if the source had duplicates, it will be output as a set of unique items.

And it will be annotated / documented accordingly too.

## Nested Models

Each attribute of a Pydantic model has a type.

But that type can itself be another Pydantic model.

So, **FastAPI** will take care of filtering out all the data that is not declared in the output model (using Pydantic).

### `response_model` or Return Type

In this case, because the two models are different, if we annotated the function return type as `UserOut`, the editor and tools would complain that we are returning an invalid type, as those are different classes.

That's why in this example we have to declare it in the `response_model` parameter.

```

...the bigger the codebase, the more noticeable it is.

Now you can create an annotated dependency once, like this:

```Python
CurrentUser = Annotated[User, Depends(get_current_user)]
```

And then you can reuse this `Annotated` dependency:

```Python
CurrentUser = Annotated[User, Depends(get_current_user)]


@app.get("/items/")
def read_items(user: CurrentUser):

   * {@code C}
   *
   * <ul>
   *   <li>All visible static methods are checked such that passing null for any parameter that's
   *       not annotated nullable (according to the rules of {@link NullPointerTester}) should throw
   *       {@link NullPointerException}.
   *   <li>If there is any visible constructor or visible static factory method declared by the

Vamos criar uma **dependência** do FastAPI com `yield` que fornecerá uma nova `Session` para cada requisição. Isso é o que garante que usamos uma única sessão por requisição. 🤓

Então, criamos uma dependência `Annotated` chamada `SessionDep` para simplificar o restante do código que usará essa dependência.

{* ../../docs_src/sql_databases/tutorial001_an_py310.py ln[25:30] hl[25:27,30] *}

### Criar Tabelas de Banco de Dados na Inicialização

   * CollectionRemoveTester.testRemove_nullAllowed, but it's not clear that it's
   * a good policy, especially because, in order for the test to pass, the
   * parameter must be misleadingly annotated as @Nullable. I suspect that
   * we'll want to remove @Nullable, add an eager checkNotNull, and loosen up
   * testRemove_nullAllowed.
   */
  @CanIgnoreReturnValue
  @Override

/**
 * A test utility that verifies that your methods and constructors throw {@link
 * NullPointerException} or {@link UnsupportedOperationException} whenever null is passed to a
 * parameter whose declaration or type isn't annotated with an annotation with the simple name
 * {@code Nullable}, {@code CheckForNull}, {@code NullableType}, or {@code NullableDecl}.
 *

We will create a FastAPI **dependency** with `yield` that will provide a new `Session` for each request. This is what ensures that we use a single session per request. 🤓

Then we create an `Annotated` dependency `SessionDep` to simplify the rest of the code that will use this dependency.

{* ../../docs_src/sql_databases/tutorial001_an_py310.py ln[25:30]  hl[25:27,30] *}

### Create Database Tables on Startup

   * {@code C}
   *
   * <ul>
   *   <li>All visible static methods are checked such that passing null for any parameter that's
   *       not annotated nullable (according to the rules of {@link NullPointerTester}) should throw
   *       {@link NullPointerException}.
   *   <li>If there is any visible constructor or visible static factory method declared by the