/**
 * {@link LoadingCache} tests that deal with empty caches.
 *
 * @author mike nonemacher
 */

@NullUnmarked
public class EmptyCachesTest extends TestCase {

  public void testEmpty() {
    for (LoadingCache<Object, Object> cache : caches()) {
      checkEmpty(cache);
    }
  }


  public void testInvalidate_empty() {
    for (LoadingCache<Object, Object> cache : caches()) {
      cache.getUnchecked("a");

/**
 * {@link LoadingCache} tests that deal with empty caches.
 *
 * @author mike nonemacher
 */

@NullUnmarked
public class EmptyCachesTest extends TestCase {

  public void testEmpty() {
    for (LoadingCache<Object, Object> cache : caches()) {
      checkEmpty(cache);
    }
  }


  public void testInvalidate_empty() {
    for (LoadingCache<Object, Object> cache : caches()) {
      cache.getUnchecked("a");

        .withKeyStrengths(ImmutableSet.of(STRONG, Strength.WEAK))
        .withValueStrengths(ImmutableSet.of(STRONG, Strength.WEAK, Strength.SOFT));
  }

  private Iterable<LoadingCache<Key, String>> caches() {
    CacheBuilderFactory factory = factoryWithAllKeyStrengths();
    return Iterables.transform(
        factory.buildAllPermutations(),
        new Function<CacheBuilder<Object, Object>, LoadingCache<Key, String>>() {

          path: |
            ~/.mimir/*.log

  consolidate-caches:
    runs-on: ${{ matrix.os }}
    strategy:
      fail-fast: false
      matrix:
        os: [ubuntu-latest, macos-latest, windows-latest]
    needs:
      - full-build
      - integration-tests
    steps:
      - name: Download Caches
        uses: actions/download-artifact@37930b1c2abaa49bbe596cd826c3c89aef350131 # v7.0.0

			remove[id] = struct{}{}
			continue
		}
	}

	// If above limit, remove the caches with the oldest handout time.
	if len(caches)-len(remove) > metacacheMaxEntries {
		remainCaches := make([]metacache, 0, len(caches)-len(remove))
		for id, cache := range caches {
			if _, ok := remove[id]; ok {
				continue
			}
			remainCaches = append(remainCaches, cache)
		}

    for (LoadingCache<Object, Object> cache : caches()) {
      // don't let the entries get GCed
      List<Entry<Object, Object>> unused = warmUp(cache);
      assertThat(cache.size()).isEqualTo(WARMUP_SIZE);
      assertMapSize(cache.asMap(), WARMUP_SIZE);
      checkValidState(cache);
    }
  }

  public void testContainsKey_found() {
    for (LoadingCache<Object, Object> cache : caches()) {
      // don't let the entries get GCed

    for (LoadingCache<Object, Object> cache : caches()) {
      // don't let the entries get GCed
      List<Entry<Object, Object>> unused = warmUp(cache);
      assertThat(cache.size()).isEqualTo(WARMUP_SIZE);
      assertMapSize(cache.asMap(), WARMUP_SIZE);
      checkValidState(cache);
    }
  }

  public void testContainsKey_found() {
    for (LoadingCache<Object, Object> cache : caches()) {
      // don't let the entries get GCed

# Tâches d'arrière-plan

Vous pouvez définir des tâches d'arrière-plan qui seront exécutées après avoir retourné une réponse.

Ceci est utile pour les opérations qui doivent avoir lieu après une requête, mais où le client n'a pas réellement besoin d'attendre que l'opération soit terminée pour recevoir une réponse.

Cela comprend, par exemple :

* Les notifications par email envoyées après l'exécution d'une action :

        .withKeyStrengths(ImmutableSet.of(STRONG, Strength.WEAK))
        .withValueStrengths(ImmutableSet.of(STRONG, Strength.WEAK, Strength.SOFT));
  }

  private Iterable<LoadingCache<Key, String>> caches() {
    CacheBuilderFactory factory = factoryWithAllKeyStrengths();
    return Iterables.transform(
        factory.buildAllPermutations(),
        new Function<CacheBuilder<Object, Object>, LoadingCache<Key, String>>() {

Wenn es überzeugende Anwendungsfälle für ein frühes Schließen bei Abhängigkeiten mit `yield` gibt, würde ich erwägen, eine neue Möglichkeit hinzuzufügen, um ein frühes Schließen optional zu aktivieren.

### Abhängigkeiten mit `yield` und `except`, Technische Details { #dependencies-with-yield-and-except-technical-details }