while (true) {
        var tryAgain = false
        var readBytesDelivered = -1L
        var errorExceptionToDeliver: IOException? = null

        // 1. Decide what to do in a synchronized block.

        withLock {
          val doReadTimeout = doReadTimeout()
          if (doReadTimeout) {
            readTimeout.enter()
          }
          try {

pkg go/doc/comment, method (*Printer) Text(*Doc) []uint8 #51082
pkg go/doc/comment, type Block interface, unexported methods #51082
pkg go/doc/comment, type Code struct #51082
pkg go/doc/comment, type Code struct, Text string #51082
pkg go/doc/comment, type Doc struct #51082
pkg go/doc/comment, type Doc struct, Content []Block #51082
pkg go/doc/comment, type Doc struct, Links []*LinkDef #51082

 *   <li><b>hash code:</b> each hash function always yields hash codes of the same fixed bit length
 *       (given by {@link #bits}). For example, {@link Hashing#sha1} produces a 160-bit number,
 *       while {@link Hashing#murmur3_32()} yields only 32 bits. Because a {@code long} value is
 *       clearly insufficient to hold all hash code values, this API represents a hash code as an
 *       instance of {@link HashCode}.

 *   <li><b>hash code:</b> each hash function always yields hash codes of the same fixed bit length
 *       (given by {@link #bits}). For example, {@link Hashing#sha1} produces a 160-bit number,
 *       while {@link Hashing#murmur3_32()} yields only 32 bits. Because a {@code long} value is
 *       clearly insufficient to hold all hash code values, this API represents a hash code as an
 *       instance of {@link HashCode}.

   * String.class}, then {@code new TypeResolver().where(formal, actual)} will {@linkplain
   * #resolveType resolve} {@code ParameterizedType List<T>} to {@code List<String>}, and resolve
   * {@code Map<T, Something>} to {@code Map<String, Something>} etc. Similarly, {@code formal} and
   * {@code actual} can be {@code Map<K, V>} and {@code Map<String, Integer>} respectively, or they

 *   <dt>From the IPv6 reserved documentation prefix ({@code 2001:db8::/32}), {@code "2001:db8::1"}.
 *   <dd>{@code 20 01 0d b8 00 00 00 00 00 00 00 00 00 00 00 01}
 *   <dt>An IPv6 "IPv4 compatible" (or "compat") address, {@code "::192.168.0.1"}.
 *   <dd>{@code 00 00 00 00 00 00 00 00 00 00 00 00 c0 a8 00 01}
 *   <dt>An IPv6 "IPv4 mapped" address, {@code "::ffff:192.168.0.1"}.
 *   <dd>{@code 00 00 00 00 00 00 00 00 00 00 ff ff c0 a8 00 01}
 * </dl>
 *

 * i.e. {@code if (key1.equals(key2))} then {@code striped.get(key1) == striped.get(key2)} (assuming
 * {@link Object#hashCode()} is correctly implemented for the keys). Note that if {@code key1} is
 * <strong>not</strong> equal to {@code key2}, it is <strong>not</strong> guaranteed that {@code
 * striped.get(key1) != striped.get(key2)}; the elements might nevertheless be mapped to the same

   *
   * <p>Java 8 introduced a new method {@code getAnnotatedBounds()} in the {@link TypeVariable}
   * interface, whose return type {@code AnnotatedType[]} is also new in Java 8. As of 2025, Android
   * has not added {@link AnnotatedType}. That means that we cannot implement that interface in
   * source code in a way that will compile on both Java and Android. If we include the {@code

Agora vamos **refatorar** este app um pouco para aumentar a **segurança** e **versatilidade**.

Se você verificar o app anterior, na interface você pode ver que, até agora, ele permite que o cliente decida o `id` do `Hero` a ser criado. 😱

    responses returned by OkHttp itself, and it was an old mistake to permit
    application code to omit a message.

 *  The challenge's scheme and realm are now non-null. If you are calling
    `new Challenge(scheme, realm)` you must provide non-null values. These were
    never null in challenges created by OkHttp, but could have been null in
    application code that creates challenges.