* and initializing on first use.
 *
 * We use this because eager classpath checks cause confusion and excessive logging in Android,
 * and we can't rely on classnames after proguard, so are probably best served by falling through
 * to a situation of trying our least likely noisiest options.
 */
class DeferredSocketAdapter(
  private val socketAdapterFactory: Factory,
) : SocketAdapter {

     * @return The toolchains model or <code>null</code> if no toolchain file was configured or the configured file does
     *         not exist.
     * @throws MisconfiguredToolchainException If the toolchain file exists but cannot be parsed.
     */
    PersistedToolchains build(File userToolchainsFile) throws MisconfiguredToolchainException;

        throw new UnsupportedOperationException();
    }

    public static List<Plugin> orderAfterMerge(
            List<Plugin> merged, List<Plugin> highPrioritySource, List<Plugin> lowPrioritySource) {
        throw new UnsupportedOperationException();
    }

        System.arraycopy(err, 0, buf, bodyStart + 8, bc);

        int decoded = resp.decode(buf, headerStart);
        assertTrue(decoded > 0);

        // When error path is taken, no blob should be parsed
        assertNull(resp.getBlob(), "Blob should be null on error path");
        assertNotNull(resp.getErrorData(), "Error data should be present");
        assertArrayEquals(err, resp.getErrorData());
    }

        testCache.releaseBuffer(originalBuffer);
        byte[] reusedBuffer = testCache.getBuffer();

        // Then
        assertSame(originalBuffer, reusedBuffer, "Released buffer should be reused");
    }

    /**
     * Simple test implementation of BufferCache for testing purposes
     */
    private static class TestBufferCacheImpl implements BufferCache {
        private final List<byte[]> cache;

on the same machine by simply switching the value of `HERMETIC_PYTHON_VERSION`
between the runs. All the python-agnostic parts of the build cache from the
previous build will be preserved and reused for the subsequent builds.

### Specifying Python dependencies

During bazel build all TensorFlow's Python dependencies are pinned to their
specific versions. This is necessary to ensure reproducibility of the build.

  private val client = OkHttpClient()

  fun run() {
    val request =
      Request
        .Builder()
        .url("http://httpbin.org/delay/2") // This URL is served with a 2 second delay.
        .build()

    val startNanos = System.nanoTime()
    val call = client.newCall(request)

    // Schedule a job to cancel the call in 1 second.
    executor.schedule({

     *
     * @param authType the type of authorization data
     * @param token the authorization data token
     * @param keys the Kerberos keys for decryption
     * @return a list of parsed authorization data
     * @throws PACDecodingException if the data cannot be decoded
     */
    public static List<KerberosAuthData> parse(int authType, byte[] token, Map<Integer, KerberosKey> keys) throws PACDecodingException {

Even though all your code is written assuming there's just `/app`.

{* ../../docs_src/behind_a_proxy/tutorial001.py hl[6] *}

        value);
    return fromIntBits(value.intValue());
  }

  /**
   * Returns an {@code UnsignedInteger} holding the value of the specified {@code String}, parsed as
   * an unsigned {@code int} value.
   *
   * @throws NumberFormatException if the string does not contain a parsable unsigned {@code int}
   *     value
   */