annotations = new ArrayList<>();
        for (Annotation a : classOrMethod.getDeclaredAnnotations()) {
          /*
           * We avoid reflecting on NullMarked because its @Target(..., MODULE) causes problems
           * under JDK 8.
           */
          if (!(a instanceof NullMarked)
              && a.annotationType().isAnnotationPresent(TesterAnnotation.class)) {
            annotations.add(a);
          }

import java.util.Set;

import org.apache.maven.artifact.Artifact;
import org.apache.maven.artifact.repository.ArtifactRepository;
import org.apache.maven.artifact.versioning.OverConstrainedVersionException;

/**
 * Specific problems during resolution that we want to account for:
 * <ul>
 *   <li>missing metadata</li>
 *   <li>version range violations</li>
 *   <li>version circular dependencies</li>
 *   <li>missing artifacts</li>

Claro, há alguns casos em que não há problema em executar as etapas anteriores várias vezes; nesse caso, é muito mais fácil de lidar.

/// tip | Dica

    @Override
    public void remove() {
      iterator.remove();
    }
  }

  public void testPeekingIteratorDoesntAdvancePrematurely() throws Exception {
    /*
     * This test will catch problems where the underlying iterator
     * throws a RuntimeException when retrieving the nth element.
     *
     * If the PeekingIterator is caching elements too aggressively,

 * https://github.com/google/guava/releases/tag/v27.0) This means that the copy of ListenableFuture
 * in com.google.guava:guava differs from the "frozen" copy in the listenablefuture artifact. This
 * could in principle cause problems for some users. Still, we expect that the benefits of the
 * nullness annotations in particular will outweigh the costs. (And it's worth noting that we have

 * https://github.com/google/guava/releases/tag/v27.0) This means that the copy of ListenableFuture
 * in com.google.guava:guava differs from the "frozen" copy in the listenablefuture artifact. This
 * could in principle cause problems for some users. Still, we expect that the benefits of the
 * nullness annotations in particular will outweigh the costs. (And it's worth noting that we have

  /**
   * Handles cases of updates involving initialization, resizing, creating new Cells, and/or
   * contention. See above for explanation. This method suffers the usual non-modularity problems of
   * optimistic retry code, relying on rechecked sets of reads.
   *
   * @param x the value
   * @param hc the hash code holder
   * @param wasUncontended false if CAS failed before call
   */

   * {@code [1]} and {@code [1, 1]}).
   */
  // Note: 90% of the time we don't add type parameters or wildcards that serve only to "tweak" the
  // desired return type. However, *nested* generics introduce a special class of problems that we
  // think tip it over into being worthwhile.
  public static <T extends @Nullable Object, S extends T> Comparator<Iterable<S>> lexicographical(
      Comparator<T> comparator) {

Y genera esquemas OpenAPI.

Así es como funciona en Flask, Starlette, Responder, etc.

Pero luego, tenemos otra vez el problema de tener una micro-sintaxis, dentro de un string de Python (un gran YAML).

If you're in a tricky situation and prepared to deal with the consequences, rewriting response headers is a powerful way to work around problems. For example, you can fix a server's misconfigured `Cache-Control` response header to enable better response caching:

```java
/** Dangerous interceptor that rewrites the server's cache-control header. */