With this patch, the external provisioner can remove the finalizer in next reconcile loop. Unfortunately if the provious existing pv has the Failed state, this patch won't take effort. It requires users to remove finalizer. ([#126043](https://github.com/kubernetes/kubernetes/pull/126043), [@carlory](https://github.com/carlory)) [SIG Apps and Storage]

  @Test
  public void testGetFinalizerUrl() {
    assertThat(getClass().getResource("internal/Finalizer.class")).isNotNull();
  }

  @Test
  public void testFinalizeClassHasNoNestedClasses() throws Exception {
    // Ensure that the Finalizer class has no nested classes.
    // See https://github.com/google/guava/issues/1505
    assertThat(Finalizer.class.getDeclaredClasses()).isEmpty();
  }

# FinalizableReferenceQueue is unused.
-keepnames class com.google.common.base.internal.Finalizer {
  *** startFinalizer(...);
}
# However, it cannot "spot" that this method needs to be kept IF the class is.
-keepclassmembers class com.google.common.base.internal.Finalizer {
  *** startFinalizer(...);
}
-keepnames class com.google.common.base.FinalizableReference {
  void finalizeReferent();
}

    assertThat(frqC).isNotSameInstanceAs(sepFrqC);
    // Check the assumptions above.

    // FRQ tries to load the Finalizer class (for the reference-collecting thread) in a few ways.
    // If the class is accessible to the system ClassLoader (ClassLoader.getSystemClassLoader())
    // then FRQ does not bother to load Finalizer.class through a separate ClassLoader. That happens

    contents = f.read()
    print(contents)
```

Por baixo dos panos, o código `open("./somefile.txt")` cria um objeto que é chamado de "Gerenciador de Contexto".

Quando o bloco `with` finaliza, ele se certifica de fechar o arquivo, mesmo que tenha ocorrido alguma exceção.

# be modified. We record the sha256 hashes of the zip files
# in fips140.sum, and the cmd/go/internal/fips140 test checks
# that the zips match.
#
# When the zip file is finalized, run 'make updatesum' to update
# fips140.sum.

default:
	@echo nothing to make

# make v1.2.3.zip builds a v1.2.3.zip file
# from the current origin/master.

and Java’s <code>volatile</code> variables.
</p>

<h3 id="finalizer">Finalizers</h3>

<p>
The <a href="/pkg/runtime/"><code>runtime</code></a> package provides
a <code>SetFinalizer</code> function that adds a finalizer to be called when
a particular object is no longer reachable by the program.

///

### `startup` e `shutdown` juntos { #startup-and-shutdown-together }

Há uma grande chance de que a lógica para sua *inicialização* e *encerramento* esteja conectada, você pode querer iniciar alguma coisa e então finalizá-la, adquirir um recurso e então liberá-lo, etc.

Fazer isso em funções separadas que não compartilham lógica ou variáveis entre si é mais difícil, pois você precisaria armazenar valores em variáveis globais ou truques semelhantes.

    * Wenn Sie FastAPI nicht verwenden und stattdessen Starlette direkt (oder ein anderes Tool wie Sanic, Flask, Responder, usw.) verwenden würden, müssten Sie die gesamte Datenvalidierung und Serialisierung selbst implementieren. Ihre finale Anwendung hätte also immer noch den gleichen Overhead, als ob sie mit FastAPI erstellt worden wäre. Und in vielen Fällen ist diese Datenvalidierung und Serialisierung der größte Teil des in Anwendungen geschriebenen Codes.

    * On n'écrit pas une application avec uniquement Uvicorn. Cela signifie que le code devrait inclure plus ou moins, au minimum, tout le code offert par Starlette (ou **FastAPI**). Et si on fait cela, l'application finale apportera les mêmes complications que si on avait utilisé un framework et que l'on avait minimisé la quantité de code et de bugs.