* explanation.
 *
 * @param <K> the most general key type this builder will be able to create caches for. This is
 *     normally {@code Object} unless it is constrained by using a method like {@link
 *     #removalListener}. Cache keys may not be null.
 * @param <V> the most general value type this builder will be able to create caches for. This is
 *     normally {@code Object} unless it is constrained by using a method like {@link

 * explanation.
 *
 * @param <K> the most general key type this builder will be able to create caches for. This is
 *     normally {@code Object} unless it is constrained by using a method like {@link
 *     #removalListener}. Cache keys may not be null.
 * @param <V> the most general value type this builder will be able to create caches for. This is
 *     normally {@code Object} unless it is constrained by using a method like {@link

          }
        }
      }
      table = newTable;
      this.count = newCount;
    }

    boolean replace(K key, int hash, V oldValue, V newValue) {
      lock();
      try {
        preWriteCleanup();

        AtomicReferenceArray<E> table = this.table;
        int index = hash & (table.length() - 1);
        E first = table.get(index);

        assertTrue(initialFile.exists(), "Initial file should exist");
        assertTrue(initialFile.isFile(), "Initial file should be a file");
        assertTrue(initialFile.canRead(), "Should be able to read initial file");

        log.info("Basic connectivity test passed - can connect, list, and read files");
    }

    @Test
    void testCreateNewFile() throws Exception {

          K[] sortedKs = (K[]) sortedKeys;
          Arrays.sort(sortedKs, comparator);
          Object[] sortedValues = new Object[size];

          // We might, somehow, be able to reorder values in-place.  But it doesn't seem like
          // there's a way around creating the separate sortedKeys array, and if we're allocating

## Version 3.0.0-RC1

_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

      .isEqualTo(Settings.DEFAULT_INITIAL_WINDOW_SIZE.toLong())
    assertThat(stream1.writeBytesMaximum)
      .isEqualTo(Settings.DEFAULT_INITIAL_WINDOW_SIZE.toLong())

    // Another stream should be able to send data even though 1 is blocked.
    val stream2 = connection.newStream(headerEntries("b", "banana"), true)
    val out2 = stream2.sink.buffer()
    out2.writeUtf8("foo")
    out2.flush()

</tr>
</tbody>
</table> </td>
</tr>
</tbody>
</table> </td>
</tr>
</tbody>
</table> </td>
</tr>
<tr>
<td colspan="2" align="left" height="12" style="height:12px;padding:0;margin:0;line-height:1px;font-size:1px;">&nbsp;</td>
</tr>
<!-- End user identity -->

    // In JDK8, table is set lazily, so it may be null.
    return table == null ? 0 : table.length;
  }

  public void testCapacityForLargeSizes() {
    int[] largeExpectedSizes =
        new int[] {
          Integer.MAX_VALUE / 2 - 1,
          Integer.MAX_VALUE / 2,
          Integer.MAX_VALUE / 2 + 1,
          Integer.MAX_VALUE - 1,
          Integer.MAX_VALUE
        };

      Table<R, C, V> table(Table<R, C, V> table, @Nullable Object mutex) {
    return new SynchronizedTable<>(table, mutex);
  }

  static final class SynchronizedTable<
          R extends @Nullable Object, C extends @Nullable Object, V extends @Nullable Object>
      extends SynchronizedObject implements Table<R, C, V> {

    SynchronizedTable(Table<R, C, V> delegate, @Nullable Object mutex) {