The problem is that GWT responds poorly to two .gwt.xml files in the same
    Java package; see https://goo.gl/pRV3Yn for details.

    The summary is that it ignores one file in favor of the other.
    util.concurrent, like nearly all our packages, has two .gwt.xml files: one
    for prod and one for tests. However, unlike our other packages, as of this
    writing it has test supersource but no prod supersource.

    The problem is that GWT responds poorly to two .gwt.xml files in the same
    Java package; see https://goo.gl/pRV3Yn for details.

    The summary is that it ignores one file in favor of the other.
    util.concurrent, like nearly all our packages, has two .gwt.xml files: one
    for prod and one for tests. However, unlike our other packages, as of this
    writing it has test supersource but no prod supersource.

  /*
   * This class maintains a lazily-initialized table of atomically
   * updated variables, plus an extra "base" field. The table size
   * is a power of two. Indexing uses masked per-thread hash codes.
   * Nearly all declarations in this class are package-private,
   * accessed directly by subclasses.
   *
   * Table entries are of class Cell; a variant of AtomicLong padded
   * to reduce cache contention on most processors. Padding is

    // up exactly as many bits as can be represented in the significand (24 for float, 53 for
    // double). That final 0 should be rounded up to 1 because the remaining bits make that number
    // slightly nearer.
    long floatConversionTest = 0xfffffe8000000002L;
    long doubleConversionTest = 0xfffffffffffff402L;

    for (long i = -3; i <= 3; i++) {
      testLongsBuilder
          .add(i)

    assertFalse(Ascii.equalsIgnoreCase("x", ""));
    assertTrue(Ascii.equalsIgnoreCase(LOWER, UPPER));
    assertTrue(Ascii.equalsIgnoreCase(UPPER, LOWER));
    // Create new strings here to avoid early-out logic.
    assertTrue(Ascii.equalsIgnoreCase(new String(IGNORED), new String(IGNORED)));
    // Compare to: "\u00c1".equalsIgnoreCase("\u00e1") == true
    assertFalse(Ascii.equalsIgnoreCase("\u00c1", "\u00e1"));

   *
   * @since 16.0
   */
  public static boolean equalsIgnoreCase(CharSequence s1, CharSequence s2) {
    // Calling length() is the null pointer check (so do it before we can exit early).
    int length = s1.length();
    if (s1 == s2) {
      return true;
    }
    if (length != s2.length()) {
      return false;
    }
    for (int i = 0; i < length; i++) {

   *
   * <p>For the case of {@link RoundingMode#HALF_EVEN}, this implementation uses the IEEE 754
   * default rounding mode: if the two nearest representable values are equally near, the one with
   * the least significant bit zero is chosen. (In such cases, both of the nearest representable
   * values are even integers; this method returns the one that is a multiple of a greater power of
   * two.)
   *

 * API</h2>
 *
 * <p>The successor to Guava's caching API is <a
 * href="https://github.com/ben-manes/caffeine/wiki">Caffeine</a>. Its API is designed to make it a
 * nearly drop-in replacement. It requires Java 8+, and is not available for Android or GWT/J2CL,
 * and may have <a href="https://github.com/ben-manes/caffeine/wiki/Guava">different (usually

 * `Function<@Nullable A, @Nullable B>` already had to find a workaround. Still, there is a *ton* of
 * fallout from trying to switch. I would be shocked if the switch would offer benefits to anywhere
 * near enough users to justify the costs.
 *
 * Fortunately, if anyone does want to use a Converter as a `Function<@Nullable A, @Nullable B>`,
 * it's easy to get one: `converter::convert`.
 *

    out.writeByte((byte) 200);
    out.writeChar('a');
    out.writeShort((short) -30000);
    out.writeShort((short) 50000);
    out.writeInt(0xCAFEBABE);
    out.writeLong(0xDEADBEEFCAFEBABEL);
    out.writeUTF("Herby Derby");
    out.writeFloat(Float.intBitsToFloat(0xCAFEBABE));
    out.writeDouble(Double.longBitsToDouble(0xDEADBEEFCAFEBABEL));

    byte[] data = baos.toByteArray();

    /* Setup input streams */