/**
   * This test attempts to create a situation in which one thread is constantly changing a file from
   * being a real directory to being a symlink to another directory. It then calls
   * deleteDirectoryContents thousands of times on a directory whose subtree contains the file
   * that's switching between directory and symlink to try to ensure that under no circumstance does

   * query than the original; your mileage may vary. Precomputation takes time and is likely to be
   * worthwhile only if the precomputed matcher is queried many thousands of times.
   *
   * <p>This method has no effect (returns {@code this}) when called in GWT: it's unclear whether a
   * precomputed matcher is faster, but it certainly consumes more memory, which doesn't seem like a

 * and memory footprint. For example, if a set of tasks are CPU-bound, one could easily create a
 * very compact {@code Striped<Lock>} of {@code availableProcessors() * 4} stripes, instead of
 * possibly thousands of locks which could be created in a {@code Map<K, Lock>} structure.
 *
 * @author Dimitris Andreou
 * @since 13.0
 */
@J2ktIncompatible
@GwtIncompatible
@ElementTypesAreNonnullByDefault

   * written into the directory.
   *
   * <p>This method assumes that the temporary volume is writable, has free inodes and free blocks,
   * and that it will not be called thousands of times per second.
   *
   * <p><b>{@link java.nio.file.Path} equivalent:</b> {@link
   * java.nio.file.Files#createTempDirectory}.
   *
   * @return the newly-created directory

   * query than the original; your mileage may vary. Precomputation takes time and is likely to be
   * worthwhile only if the precomputed matcher is queried many thousands of times.
   *
   * <p>This method has no effect (returns {@code this}) when called in GWT: it's unclear whether a
   * precomputed matcher is faster, but it certainly consumes more memory, which doesn't seem like a