}
    }
    return string;
  }

  /**
   * Returns the escaped form of the given character, or {@code null} if this character does not
   * need to be escaped. If an empty array is returned, this effectively strips the input character
   * from the resulting text.
   *
   * <p>If the character does not need to be escaped, this method should return {@code null}, rather

       * by stripping annotated methods entirely, and if we strip this method, then JUnit would just
       * run the supermethod as usual.
       *
       * TODO: b/292578973: Use @AndroidIncompatible if we change our system to keep the methods in
       * place but to have the test runner skip them. However, note that if we choose to *both*
       * strip the methods *and* have the test runner not run them (for some unusual cases in which

                  </fileset>
                </copy>
                <!-- TODO(cpovirk): If we continue stripping these once we use @ElementTypesAreNonnullByDefault, then we should strip @ElementTypesAreNonnullByDefault, too, to reflect that the classes are *not* fully annotated. But hopefully we can stop stripping them, as https://github.com/jspecify/jspecify/issues/184 should make possible. -->

 *       that we could probably work around. Or we could manually strip the test from open-source
 *       Guava while continuing to run it internally, as we do with many other tests. This would
 *       suffice because we our Android users and tests are using the open-source version, which
 *       would no longer have the problematic test. But why bother when we can instead strip it with
 *       a more precisely named annotation?

   * PackageSanityTests. (The test would run on the JVM, too, if not for the suppression below, and
   * that would be a problem because it violates small-test rules. Note that we strip the
   * suppression externally, but it's OK because we don't enforce test-size rules there.)
   *
   * We'd just use PackageSanityTests directly, saving us from needing this separate type, but we're

   * PackageSanityTests. (The test would run on the JVM, too, if not for the suppression below, and
   * that would be a problem because it violates small-test rules. Note that we strip the
   * suppression externally, but it's OK because we don't enforce test-size rules there.)
   *
   * We'd just use PackageSanityTests directly, saving us from needing this separate type, but we're

  The Dormouse had closed its eyes by this time, and was going
off into a doze; but, on being pinched by the Hatter, it woke up
again with a little shriek, and went on:  `--that begins with an
M, such as mouse-traps, and the moon, and memory, and muchness--
you know you say things are "much of a muchness"--did you ever
see such a thing as a drawing of a muchness?'

  `Really, now you ask me,' said Alice, very much confused, `I

  private Striped() {}

  /**
   * Returns the stripe that corresponds to the passed key. It is always guaranteed that if {@code
   * key1.equals(key2)}, then {@code get(key1) == get(key2)}.
   *
   * @param key an arbitrary, non-null key
   * @return the stripe that the passed key corresponds to
   */
  public abstract L get(Object key);

  /**

      value: >
        Comparing two approaches to a use case side by side can make it easier to examine the
        differences between them.


        Additionally, it's very useful to us if you can provide a "straw API" — what the
        method signatures would look like, for example, even if the method and class names are still
        in flux. This can make the feature you're suggesting much clearer to us.

  - type: textarea

   * is guaranteed to be in the range {@code 0 <= cp <= Character#MAX_CODE_POINT}.
   *
   * <p>If an empty array is returned, this effectively strips the input character from the
   * resulting text.
   *
   * <p>If the character does not need to be escaped, this method should return {@code null}, rather