} catch (final IOException e) {
                    throw new StorageException("Failed to load GCS credentials from " + credentialsPath, e);
                }
            }
            // If no credentials path, uses default credentials (GOOGLE_APPLICATION_CREDENTIALS env var)
        }

        this.storage = builder.build().getService();
    }

    @Override

/**
 * RegexRule is a rule implementation that uses regular expressions to match against the ResponseData.
 * It allows defining multiple regular expressions for different fields of the ResponseData.
 * The rule can be configured to require all regular expressions to match (allRequired = true) or
 * only one of them (allRequired = false). It also supports a default rule that always matches.
 *
 * <p>

 * This class provides methods to store, retrieve, update, and delete access results,
 * as well as to iterate over them. It uses a {@link MemoryDataHelper} to store the data in memory.
 *
 * <p>
 * The class uses a static {@code idCount} to generate unique IDs for each access result.
 * The {@code idCountLock} object is used to synchronize access to the {@code idCount} variable,

     * Tests a filter that uses directory path information.
     */
    @Test
    @DisplayName("filter can use SmbFile directory for decision logic")
    void filterCanUseSmbFileForDecision() throws Exception {
        // Stub the getPath method to return a specific value
        when(mockDir.getPath()).thenReturn("/share/folder/");

        // Implementation uses dir.getPath() to check if file should be in that path

# /---------------------------------------------------------------------------
# classificationDeploymentMap: (NotRequired - Default map:{})
#
# The relation between column and classification.
#
# This property uses classification names of classificationDefinitionMap.
# The table name '$$ALL$$' means all tables are target.
# The table names and column names are treated as case insensitive.
#
# You don't need specify here about table classifications.

handshake. The best part of this example is that we don't need to make our test code insecure with a
a fake `HostnameVerifier` or `X509TrustManager`.

Certificate Authorities
-----------------------

The above example uses a self-signed certificate. This is convenient for testing but not
representative of real-world HTTPS deployment. To get closer to that we can use `HeldCertificate`

    for (E element : collection) { // uses iterator()
      iteratorElements.add(element);
    }
    assertEqualIgnoringOrder(asList(createSamplesArray()), iteratorElements);
  }

  @CollectionFeature.Require(KNOWN_ORDER)
  public void testIterationOrdering() {
    List<E> iteratorElements = new ArrayList<>();
    for (E element : collection) { // uses iterator()
      iteratorElements.add(element);
    }

call.executeAsync().use { response ->
  withContext(Dispatchers.IO) {
    println(response.body?.string())
  }
}
```

This is implemented using `suspendCancellableCoroutine`
but uses the standard Dispatcher in OkHttp. This means
that by default Kotlin's Dispatchers are not used.

Cancellation if implemented sensibly in both directions.
Cancelling a coroutine scope, will cancel the call.

  //   timeout is small we shouldn't park(). This needs to be traded off with the cpu overhead of
  //   spinning, so we use SPIN_THRESHOLD_NANOS which is what AbstractQueuedSynchronizer uses for
  //   similar purposes.
  // * We want to behave reasonably for timeouts of 0
  // * We are more responsive to completion than timeouts. This is because parkNanos depends on

/**
 * Implementation of the {@link UrlQueueService} interface.
 * This class provides methods for managing a queue of URLs to be crawled,
 * including adding, deleting, and retrieving URLs from the queue.
 * It uses a {@link MemoryDataHelper} to store the URL queue data in memory.
 *
 * <p>
 * The class is responsible for:
 * </p>
 * <ul>
 *   <li>Updating session IDs for URL queues.</li>