for k, v := range subset.Annotations {
			if k == "proxy.istio.io/config" {
				if err := patchProxyConfigFile(path.Join(subsetDir, "mesh.yaml"), v); err != nil {
					return fmt.Errorf("failed patching proxyconfig: %v", err)
				}
			}
		}

		if err := customizeVMEnvironment(ctx, cfg, path.Join(subsetDir, "cluster.env"), istiodAddr); err != nil {
			return fmt.Errorf("failed customizing cluster.env: %v", err)

#### Caching

Config translation typically takes the overwhelming majority of Istiod's resource usage. In particular, protobuf encoding. As a result, caching has been introduced, storing the already encoded `protobuf.Any` for a given resource.

			// if clusterIPs are the same, then clear on their behalf
			if sameClusterIPs(oldSvc, newSvc) {
				newSvc.Spec.ClusterIPs = nil
			}

			// if they provided nil, then we are fine (handled by patching case above)
			// if they changed it then validation will catch it
		} else {
			// ClusterIP has changed but not cleared *and* ClusterIPs are the same
			// then we set ClusterIPs based on ClusterIP

	updatePod *example.Pod

	// expectedPod is the pod that you expect to get back after the patch is complete
	expectedPod   *example.Pod
	expectedError string
	// if set, indicates the number of times patching was expected to be attempted
	expectedTries int
}

func (tc *patchTestCase) Run(t *testing.T) {
	t.Logf("Starting test %s", tc.name)

	namespace := tc.startingPod.Namespace
	name := tc.startingPod.Name

	}
	baseJSON, err := base.JSON()
	if err != nil {
		return nil, err
	}

	// get a versioned object from the scheme, we can use the strategic patching mechanism
	// (i.e. take advantage of patchStrategy in the type)
	versionedObject, err := scheme.Scheme.New(base.GroupVersionKind())
	if err != nil {
		return nil, err
	}
	// strategic merge patch

			virtualHosts = append(virtualHosts, v)
		}
	}

	util.SortVirtualHosts(virtualHosts)

	routeCfg := &route.RouteConfiguration{
		// Retain the routeName as its used by EnvoyFilter patching logic
		Name:                           routeName,
		VirtualHosts:                   virtualHosts,
		ValidateClusters:               proto.BoolFalse,
		IgnorePortInHostMatching:       !node.IsProxylessGrpc(),

For example, the `javaCompile` task action calls the Java compiler to transform source code into byte code.

It is possible to dynamically modify task actions for tasks that are already registered.
This is helpful for testing, patching, or modifying core build logic.

Let's look at an example of a simple Gradle build with one `app` subproject that makes up a Java application – containing one Java class and using Gradle's `application` plugin.

        and:
        mavenHttpRepo.getModuleMetaData("org.test", "projectA").expectGet()
        matchingA.pom.expectGet()
        matchingA.artifact.expectGet()

        mavenHttpRepo.getModuleMetaData("org.test", "projectB").expectGet()
        matchingB.metaData.expectGet()
        matchingB.pom.expectGet()
        matchingB.artifact.expectGet()

        and:
        run 'retrieve'

        then:

        when:
        1 * one.missing >> false
        0 * componentSelectionRules.apply(_, _)

        then:
        chooser.selectNewestComponent(one, two) == one
    }

    def "chooses newest matching version without requiring metadata"() {
        given:
        def a = component('1.2')
        def b = component('1.3')
        def c = component('2.0')

        executedAndNotSkipped(":outer")
    }

    def "fails when using continuous build and watching is disabled"() {
        buildFile << """
            task myTask
        """

        expect:
        fails("myTask", "--no-watch-fs")
        failureDescriptionContains("Continuous build does not work when file system watching is disabled")
    }

    private void exitContinuousBuildSinceNoDeclaredInputs() {