tasks.withType(Test.class).configureEach(task -> extension.fixtures.all(fixtureProject -> {
            task.dependsOn(fixtureProject.getTasks().named("postProcessFixture"));
            task.finalizedBy(fixtureProject.getTasks().named("composeDown"));
            configureServiceInfoForTask(
                task,
                fixtureProject,
                true,

                    throw new DictionaryException("Failed to write: " + item, e);
                }
            }
            return null;
        }

        /**
         * Closes the updater and finalizes the file update operation.
         * If changes were committed, the temporary file replaces the original.
         * Otherwise, the temporary file is deleted.
         */
        @Override
        public void close() {

                }
            }
        }

        return this;
    }

    /**
     * {@inheritDoc}
     *
     * @see java.lang.Object#finalize()
     */
    @Override
    protected void finalize() throws Throwable {
        try {
            if (isConnected() && this.usageCount.get() != 0) {

///

### `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.

          "Unable to cast input to GraphTensor");
    }
    inputs_.push_back(t->output_);
    *output = tensorflow::down_cast<TracingTensorHandle*>(outputs[0]);
    return absl::OkStatus();
  }

  absl::Status Finalize(OutputList* outputs, AbstractFunction** f) override {
    std::vector<TF_Output> graph_outputs;
    graph_outputs.reserve(outputs->outputs.size());
    for (auto* abstract_output : outputs->outputs) {

	// profitable. To achieve this we create a goroutine per-proc.
	// These goroutines access considerable amount of local data so that
	// unnecessary rescheduling is penalized by cache misses.
	m := NewDRWMutex(ds, "")
	var acc0, acc1 uint64
	b.RunParallel(func(pb *testing.PB) {
		var data [16 << 10]uint64
		for i := 0; pb.Next(); i++ {
			m.Lock(id, source)
			acc0 -= 100

    // Extract the resulting tensor.
    add_output2 = TF_OutputListGet(add_outputs, 0);
    TF_DeleteOutputList(add_outputs);
  }

  TF_DeleteAbstractTensor(arg0);
  TF_DeleteAbstractTensor(arg1);

  // Finalize the function by providing the returned values.
  TF_AbstractFunction* func;
  {
    // We want to return the output of both add operations, create a new list
    // and populate it.

O benefício de lançar uma exceção em vez de retornar um valor ficará mais evidente na seção sobre Dependências e Segurança.

 * about crawler operations including timing data, performance metrics, and
 * operational events. It uses an internal cache to maintain statistics
 * objects and provides methods to begin tracking, record events, and
 * finalize statistics collection.
 *
 */
public class CrawlerStatsHelper {

    /**
     * Creates a new instance of CrawlerStatsHelper.
     */
    public CrawlerStatsHelper() {
        // Default constructor
    }

      "coverage" : {
        "per_commit" : [ "linux", "windows", "macOs" ]
      }
    } ]
  }, {
    "testId" : "org.gradle.performance.regression.android.RealLifeAndroidBuildPerformanceTest.calculate task graph with test finalizer",
    "groups" : [ {
      "testProject" : "largeAndroidBuild",
      "coverage" : {
        "per_commit" : [ "linux" ]
      }
    } ]
  }, {