// flow-control window. This is necessary because we don't transmit window updates until
          // the application reads the data. If we throw this prior to updating the connection
          // flow-control window, we risk having it go to 0 preventing the server from sending data.
          throw errorExceptionToDeliver!!
        }

        return -1L // This source is exhausted.
      }
    }

      }
    }
    return args;
  }

  private <T> @Nullable T getDefaultValue(TypeToken<T> type) {
    // We assume that all defaults are generics-safe, even if they aren't,
    // we take the risk.
    @SuppressWarnings("unchecked")
    T defaultValue = (T) defaults.getInstance(type.getRawType());
    if (defaultValue != null) {
      return defaultValue;
    }

       *   user task -- which means that we already got past this code (or else we exited early
       *   above))
       */
      // Unconditionally set; there is no risk of throwing away a queued task from another thread,
      // because in order for the current task to run on this executor the previous task must have

        // The BigInteger implementation is tested separately, use it as the reference.
        // Promote the int value (rather than using intValue() on the expected value) to avoid
        // any risk of truncation which could lead to a false positive.
        assertEquals(BigIntegerMath.sqrt(valueOf(x), mode), valueOf(IntMath.sqrt(x, mode)));
      }
    }
  }

   * </ul>
   *
   * <p>A "coerced" IPv4 address is equivalent to itself.
   *
   * <p>NOTE: This method is failsafe for security purposes: ALL IPv6 addresses (except localhost
   * (::1)) are hashed to avoid the security risk associated with extracting an embedded IPv4
   * address that might permit elevated privileges.
   *
   * @param ip {@link InetAddress} to "coerce"
   * @return {@link Inet4Address} represented "coerced" address

_2016-01-02_

OkHttp 3 is a major release focused on API simplicity and consistency. The API
changes are numerous but most are cosmetic. Applications should be able to
upgrade from the 2.x API to the 3.x API mechanically and without risk.

Because the release includes breaking API changes, we're changing the project's
package name from `com.squareup.okhttp` to `okhttp3`. This should make it

    for (long x : POSITIVE_LONG_CANDIDATES) {
      for (RoundingMode mode : ALL_SAFE_ROUNDING_MODES) {
        // Promote the long value (rather than using longValue() on the expected value) to avoid
        // any risk of truncation which could lead to a false positive.
        assertEquals(BigIntegerMath.sqrt(valueOf(x), mode), valueOf(LongMath.sqrt(x, mode)));
      }
    }
  }

   * advantage of <a
   * href="https://docs.oracle.com/javase/tutorial/java/generics/genTypeInference.html#type-inference-instantiation">"diamond"
   * syntax</a>. (Unlike here, there is no risk of overload ambiguity, since the {@code ArrayList}
   * constructors very wisely did not accept varargs.)
   *
   * @param initialArraySize the exact size of the initial backing array for the returned array list

   * EntryTransformer} key type {@code K}, {@code k.equals(k2)} implies that {@code k2} is also of
   * type {@code K}. Using an {@code EntryTransformer} key type for which this may not hold, such as
   * {@code ArrayList}, may risk a {@code ClassCastException} when calling methods on the
   * transformed multimap.
   *
   * @since 7.0
   */
  public static <

   * EntryTransformer} key type {@code K}, {@code k.equals(k2)} implies that {@code k2} is also of
   * type {@code K}. Using an {@code EntryTransformer} key type for which this may not hold, such as
   * {@code ArrayList}, may risk a {@code ClassCastException} when calling methods on the
   * transformed multimap.
   *
   * @since 7.0
   */
  public static <