// ----------------------------------------------------------------------------
        // I'm fully aware that the file could just be moved using File.rename but
        // there are bugs in various JVM that have problems doing this across
        // different filesystem. So we'll incur the small hit to actually copy
        // here and be safe. jvz.
        // ----------------------------------------------------------------------------

This has been mostly done, but there remain a few holdouts (fixing these is out of scope).

To keep the Groovy DSL ergonomic, we generate methods as necessary from the non-Groovy equivalents.

Doing this provides the following specific benefits:
- **Reduce the API surface** - We no longer need to maintain two methods.
- **Consistency** - All languages have consistent access to the same APIs and ergonomics in the DSL.

    @Test
    void testCalculationOfBuildPlanWithIndividualTaskWherePluginIsSpecifiedInThePom() throws Exception {
        // We are doing something like "mvn resources:resources" where no version is specified but this
        // project we are working on has the version specified in the POM so the version should come from there.

/**
 * Extracts Kotlin DSL runtime generated sources.
 *
 * Current implementation extracts these from the wrapper's API jars.
 * This is not correct as it should do this with the built distribution instead.
 *
 * Doing it correctly would require running a Gradle build with the full
 * distribution and extracting the generated api jar from its Gradle user home,
 * slowing down building documentation.
 *

#### Optimizations

A naive implementation would simply regenerate all resources, of all subscribed types, for each client, on any configuration change. However, this scales poorly. As a result, we have many levels of optimizations to avoid doing this work.

      <version>4.13.2</version>
    </dependency>
    <dependency>
      <!--
      Do not include Truth in non-test scope! Doing so creates a problematic dependency cycle.
      -->
      <groupId>com.google.truth</groupId>
      <artifactId>truth</artifactId>
      <version>${truth.version}</version>
      <scope>test</scope>
      <exclusions>

      for (int j = 0; j < datasetSize; j++) {
        datasets[i][j] = rng.nextDouble();
      }
    }
  }

  private double[] dataset(int i) {
    // We must test on a fresh clone of the dataset each time. Doing sorts and quickselects on a
    // dataset which is already sorted or partially sorted is cheating.
    return datasets[i & 0xFF].clone();
  }

  @Benchmark
  double median(int reps) {
    double dummy = 0.0;

      for (int j = 0; j < datasetSize; j++) {
        datasets[i][j] = rng.nextDouble();
      }
    }
  }

  private double[] dataset(int i) {
    // We must test on a fresh clone of the dataset each time. Doing sorts and quickselects on a
    // dataset which is already sorted or partially sorted is cheating.
    return datasets[i & 0xFF].clone();
  }

  @Benchmark
  double median(int reps) {
    double dummy = 0.0;

# Using the Request Directly

Up to now, you have been declaring the parts of the request that you need with their types.

Taking data from:

* The path as parameters.
* Headers.
* Cookies.
* etc.

And by doing so, **FastAPI** is validating that data, converting it and generating documentation for your API automatically.

But there are situations where you might need to access the `Request` object directly.

import org.junit.jupiter.api.Tag
import org.junit.jupiter.api.Test

/**
 * Integration test to confirm that [TaskRunner] works with a real backend. Business logic is all
 * exercised by [TaskRunnerTest].
 *
 * This test is doing real sleeping with tolerances of 250 ms. Hopefully that's enough for even the
 * busiest of CI servers.
 */
@Tag("Slowish")
class TaskRunnerRealBackendTest {
  private val log = LinkedBlockingDeque<String>()