```

This results in the following execution order:

* **Request**: MiddlewareB → MiddlewareA → route

* **Response**: route → MiddlewareA → MiddlewareB

This stacking behavior ensures that middlewares are executed in a predictable and controllable order.

## Other middlewares { #other-middlewares }

      object : Task(name) {
        override fun runOnce(): Long = block()
      },
      delayNanos,
    )
  }

  /**
   * Executes [block] once on a task runner thread.
   *
   * TODO: make this inline once this is fixed: https://github.com/oracle/graal/issues/3466
   */
  fun execute(
    name: String,
    delayNanos: Long = 0L,
    cancelable: Boolean = true,
    block: () -> Unit,
  ) {
    schedule(

 *
 * <p>Common packaging types include {@code jar}, {@code war}, {@code pom}, {@code maven-plugin}, {@code ear}, and others.
 * These types influence various aspects of the build lifecycle, such as which plugins are executed and how dependencies are managed.</p>
 *
 * <p>The {@code Packaging} class is an immutable value object, ensuring that once a packaging type is defined, it cannot be changed.</p>
 *
 * <h2>Standard Packaging Types</h2>

 * {@link #shutDown} and also a {@link #runOneIteration} method that will be executed periodically.
 *
 * <p>This class uses the {@link ScheduledExecutorService} returned from {@link #executor} to run
 * the {@link #startUp} and {@link #shutDown} methods and also uses that service to schedule the
 * {@link #runOneIteration} that will be executed periodically as specified by its {@link

Calls are executed in one of two ways:

 * **Synchronous:** your thread blocks until the response is readable.

        // Default constructor
    }

    /**
     * Executes the query command and returns a QueryBuilder.
     * @param context The query context containing search parameters.
     * @param query The Lucene query to execute.
     * @param boost The boost factor to apply.
     * @return The executed QueryBuilder.
     */
    public abstract QueryBuilder execute(final QueryContext context, final Query query, final float boost);

programs that TensorFlow executes. TensorFlow programs are encoded as
computation
[**graphs**](https://developers.google.com/machine-learning/glossary/#graph).
Since models are practically programs that TensorFlow executes, using untrusted
models or graphs is equivalent to running untrusted code.

If you need to run untrusted models, execute them inside a

     */
    @Nonnull
    Map<String, String> systemProperties();

    /**
     * Returns the top-level directory of the Maven invocation.
     * This is typically the directory containing the POM file being executed.
     *
     * @return the top-level directory path
     */
    @Nonnull
    Path topDirectory();

    /**

 * {@link #shutDown} and also a {@link #runOneIteration} method that will be executed periodically.
 *
 * <p>This class uses the {@link ScheduledExecutorService} returned from {@link #executor} to run
 * the {@link #startUp} and {@link #shutDown} methods and also uses that service to schedule the
 * {@link #runOneIteration} that will be executed periodically as specified by its {@link

    }

    /**
     * Executes an external command with the given parameters.
     * @param commands The command array to execute.
     * @param username The username parameter for the command.
     * @param password The password parameter for the command.
     * @return The exit code of the executed command.
     */