* href="https://url.spec.whatwg.org/#syntax-url-path-segment">URL path segments</a>. The returned
   * escaper escapes all non-ASCII characters, even though <a
   * href="https://url.spec.whatwg.org/#url-code-points">many of these are accepted in modern
   * URLs</a>. (<a href="https://url.spec.whatwg.org/#path-state">If the escaper were to leave these
   * characters unescaped, they would be escaped by the consumer at parse time, anyway.</a>)

   * only the entries for {@code keys} will be returned from {@code getAll}.
   *
   * <p>This method should be overridden when bulk retrieval is significantly more efficient than
   * many individual lookups. Note that {@link LoadingCache#getAll} will defer to individual calls
   * to {@link LoadingCache#get} if this method is not overridden.
   *

     * best we can do (since logging may overflow the stack) is to let the error propagate. Because
     * it is an Error, it won't be caught and logged by AbstractFuture.executeListener. Instead, it
     * can propagate through many layers of AbstractTransformFuture up to the root call to set().
     *
     * https://github.com/google/guava/issues/2254
     *
     * Other kinds of Errors are possible:
     *

```java
/** This interceptor compresses the HTTP request body. Many webservers can't handle this! */
final class GzipRequestInterceptor implements Interceptor {
  @Override public Response intercept(Interceptor.Chain chain) throws IOException {
    Request originalRequest = chain.request();

    expected.addAll(index, elements);
    expectContents(expected);
  }

  /*
   * TODO: if we're testing a list, we could check indexOf(). (Doing it in
   * AbstractListTester isn't enough because many tests that run on lists don't
   * extends AbstractListTester.) We could also iterate over all elements to
   * verify absence
   */
  protected void expectMissing(E... elements) {
    for (E element : elements) {

        /*
         * It's very unlikely that any loaded Futures class will see getChecked called with more
         * than a handful of exceptions. But it seems prudent to set a cap on how many we'll cache.
         * This avoids out-of-control memory consumption, and it keeps the cache from growing so
         * large that doing the lookup is noticeably slower than redoing the work would be.
         *

   * that satisfy the filter will be removed from the underlying set.
   *
   * <p>The returned set isn't threadsafe or serializable, even if {@code unfiltered} is.
   *
   * <p>Many of the filtered set's methods, such as {@code size()}, iterate across every element in
   * the underlying set and determine which elements satisfy the filter. When a live view is

      fun run() {
        val request = Request.Builder()
            .url("http://publicobject.com/secrets/hellosecret.txt")
            .build()
      }
    ```

    To avoid making many retries when authentication isn't working, you can return null to give up. For example, you may want to skip the retry when these exact credentials have already been attempted:

    ```kotlin

internal fun FileSystem.deleteIfExists(path: Path) {
  try {
    delete(path)
  } catch (fnfe: FileNotFoundException) {
    return
  }
}

/** Tolerant delete, try to clear as many files as possible even after a failure. */
internal fun FileSystem.deleteContents(directory: Path) {
  var exception: IOException? = null
  val files =
    try {
      list(directory)

HTTPS
=====

OkHttp attempts to balance two competing concerns:

 * **Connectivity** to as many hosts as possible. That includes advanced hosts that run the latest versions of [boringssl](https://boringssl.googlesource.com/boringssl/) and less out of date hosts running older versions of [OpenSSL](https://www.openssl.org/).