*
   * @throws NullPointerException if {@code elements} contains a null element
   */
  public static <E> ImmutableList<E> copyOf(Iterator<? extends E> elements) {
    // We special-case for 0 or 1 elements, but going further is madness.
    if (!elements.hasNext()) {
      return of();
    }
    E first = elements.next();
    if (!elements.hasNext()) {
      return of(first);
    } else {

     *   attempt multiple proxy servers in sequence, eventually falling back to a direct
     *   connection.
     *
     * Set this to false to avoid retrying requests when doing so is destructive. In this case the
     * calling application should do its own recovery of connectivity failures.
     */
    fun retryOnConnectionFailure(retryOnConnectionFailure: Boolean) =
      apply {

      .isEqualTo(bytes)
    assertThat(CertificateAdapters.extension.fromDer(bytes))
      .isEqualTo(extension)
  }

  /** Tags larger than 30 are a special case. */
  @Test fun `large tag`() {
    val bytes = "df83fb6800".decodeHex()

    val adapter = Adapters.NULL.withTag(tagClass = DerHeader.TAG_CLASS_PRIVATE, tag = 65_000L)
    assertThat(adapter.toDer(null)).isEqualTo(bytes)

### Other notable changes

* Kubernetes json deserializer is now case-sensitive to restore compatibility with pre-1.8 servers. ([#65157](https://github.com/kubernetes/kubernetes/pull/65157), [@caesarxuchao](https://github.com/caesarxuchao))
    * If your config files contains fields with wrong case, the config files will be now invalid.

   *
   * <p><b>Exception:</b> for consistency with separator-based splitters, {@code split("")} does not
   * yield an empty iterable, but an iterable containing {@code ""}. This is the only case in which
   * {@code Iterables.size(split(input))} does not equal {@code IntMath.divide(input.length(),
   * length, CEILING)}. To avoid this behavior, use {@code omitEmptyStrings}.
   *

    return from(Arrays.asList(elements));
  }

  /**
   * Construct a fluent iterable from another fluent iterable. This is obviously never necessary,
   * but is intended to help call out cases where one migration from {@code Iterable} to {@code
   * FluentIterable} has obviated the need to explicitly convert to a {@code FluentIterable}.
   *

    // array slot is read and written exactly once. However, it can have very poor memory locality:
    // benchmarking shows it can take 7 times longer than the other two in some cases. The other two
    // do n swaps, minus a delta (0 or 2 for Reversal, gcd(d, n) for Successive), so that's about
    // twice as many reads and writes. But benchmarking shows that they usually perform better than

apiserver to same-host redirects.  If nodes are configured to respond to CRI streaming requests on a different host interface than what the apiserver makes requests on (only the case if not using the built-in dockershim & setting the kubelet flag `--redirect-container-streaming=true`), then these requests will be broken. In that case, the feature can be temporarily disabled until the node configuration is corrected. We suggest setting `--redirect-container-streaming=false` on the kubelet to avoid issue...

        ptr[i] = strs[i].c_str();
      }
    }
    return ptr;
  }

  // An explicit `num_opers` is needed so that we can distinguish between the
  // case of no operations specified (-1) and the case of an empty set of
  // operations specified (0).
  void DefineT(int num_opers, const std::vector<TF_Operation*>& opers,
               const std::vector<TF_Output>& inputs,

validation. Bypassing this validation could allow authenticated requests destined for Nodes to be redirected to the API Server through its private network.

The merged fix enforces validation against the proxying address for a Node. In some cases, the fix can break clients that depend on the `nodes/proxy` subresource, specifically if a kubelet advertises a localhost or link-local address to the Kubernetes control plane. Configuring an egress proxy for egress to the cluster network can also...