Testing Build Logic with TestKit
The Gradle TestKit (a.k.a. just TestKit) is a library that aids in testing Gradle plugins and build logic generally. At this time, it is focused on functional testing. That is, testing build logic by exercising it as part of a programmatically executed build. Over time, the TestKit will likely expand to facilitate other kinds of tests.
Usage
To use the TestKit, include the following in your plugin’s build:
dependencies {
testImplementation gradleTestKit()
}
dependencies {
testImplementation(gradleTestKit())
}
The gradleTestKit()
encompasses the classes of the TestKit, as well as the Gradle Tooling API client. It does not include a version of JUnit, TestNG, or any other test execution framework. Such a dependency must be explicitly declared.
dependencies {
testImplementation("org.junit.jupiter:junit-jupiter:5.7.1")
}
tasks.withType(Test).configureEach {
useJUnitPlatform()
}
dependencies {
testImplementation("org.junit.jupiter:junit-jupiter:5.7.1")
}
tasks.withType<Test>().configureEach {
useJUnitPlatform()
}
Functional testing with the Gradle runner
The GradleRunner facilitates programmatically executing Gradle builds, and inspecting the result.
A contrived build can be created (e.g. programmatically, or from a template) that exercises the “logic under test”. The build can then be executed, potentially in a variety of ways (e.g. different combinations of tasks and arguments). The correctness of the logic can then be verified by asserting the following, potentially in combination:
-
The build’s output;
-
The build’s logging (i.e. console output);
-
The set of tasks executed by the build and their results (e.g. FAILED, UP-TO-DATE etc.).
After creating and configuring a runner instance, the build can be executed via the GradleRunner.build() or GradleRunner.buildAndFail() methods depending on the anticipated outcome.
The following demonstrates the usage of the Gradle runner in a Java JUnit test:
Example: Using GradleRunner with Java and JUnit
import org.gradle.testkit.runner.BuildResult;
import org.gradle.testkit.runner.GradleRunner;
import org.junit.jupiter.api.BeforeEach;
import org.junit.jupiter.api.Test;
import org.junit.jupiter.api.io.TempDir;
import java.io.BufferedWriter;
import java.io.File;
import java.io.FileWriter;
import java.io.IOException;
import static org.gradle.testkit.runner.TaskOutcome.SUCCESS;
import static org.junit.jupiter.api.Assertions.assertEquals;
import static org.junit.jupiter.api.Assertions.assertTrue;
public class BuildLogicFunctionalTest {
@TempDir File testProjectDir;
private File settingsFile;
private File buildFile;
@BeforeEach
public void setup() {
settingsFile = new File(testProjectDir, "settings.gradle");
buildFile = new File(testProjectDir, "build.gradle");
}
@Test
public void testHelloWorldTask() throws IOException {
writeFile(settingsFile, "rootProject.name = 'hello-world'");
String buildFileContent = "task helloWorld {" +
" doLast {" +
" println 'Hello world!'" +
" }" +
"}";
writeFile(buildFile, buildFileContent);
BuildResult result = GradleRunner.create()
.withProjectDir(testProjectDir)
.withArguments("helloWorld")
.build();
assertTrue(result.getOutput().contains("Hello world!"));
assertEquals(SUCCESS, result.task(":helloWorld").getOutcome());
}
private void writeFile(File destination, String content) throws IOException {
BufferedWriter output = null;
try {
output = new BufferedWriter(new FileWriter(destination));
output.write(content);
} finally {
if (output != null) {
output.close();
}
}
}
}
Any test execution framework can be used.
As Gradle build scripts can also be written in the Groovy programming language, it is often a productive choice to write Gradle functional tests in Groovy. Furthermore, it is recommended to use the (Groovy based) Spock test execution framework as it offers many compelling features over the use of JUnit.
The following demonstrates the usage of the Gradle runner in a Groovy Spock test:
Example: Using GradleRunner with Groovy and Spock
import org.gradle.testkit.runner.GradleRunner
import static org.gradle.testkit.runner.TaskOutcome.*
import spock.lang.TempDir
import spock.lang.Specification
class BuildLogicFunctionalTest extends Specification {
@TempDir File testProjectDir
File settingsFile
File buildFile
def setup() {
settingsFile = new File(testProjectDir, 'settings.gradle')
buildFile = new File(testProjectDir, 'build.gradle')
}
def "hello world task prints hello world"() {
given:
settingsFile << "rootProject.name = 'hello-world'"
buildFile << """
task helloWorld {
doLast {
println 'Hello world!'
}
}
"""
when:
def result = GradleRunner.create()
.withProjectDir(testProjectDir)
.withArguments('helloWorld')
.build()
then:
result.output.contains('Hello world!')
result.task(":helloWorld").outcome == SUCCESS
}
}
It is a common practice to implement any custom build logic (like plugins and task types) that is more complex in nature as external classes in a standalone project. The main driver behind this approach is bundle the compiled code into a JAR file, publish it to a binary repository and reuse it across various projects.
Getting the plugin-under-test into the test build
The GradleRunner uses the Tooling API to execute builds. An implication of this is that the builds are executed in a separate process (i.e. not the same process executing the tests). Therefore, the test build does not share the same classpath or classloaders as the test process and the code under test is not implicitly available to the test build.
Starting with version 2.13, Gradle provides a conventional mechanism to inject the code under test into the test build.
Automatic injection with the Java Gradle Plugin Development plugin
The Java Gradle Plugin development plugin can be used to assist in the development of Gradle plugins.
Starting with Gradle version 2.13, the plugin provides a direct integration with TestKit.
When applied to a project, the plugin automatically adds the gradleTestKit()
dependency to the testApi
configuration.
Furthermore, it automatically generates the classpath for the code under test and injects it via GradleRunner.withPluginClasspath() for any GradleRunner
instance created by the user.
It’s important to note that the mechanism currently only works if the plugin under test is applied using the plugins DSL.
If the target Gradle version is prior to 2.8, automatic plugin classpath injection is not performed.
The plugin uses the following conventions for applying the TestKit dependency and injecting the classpath:
-
Source set containing code under test:
sourceSets.main
-
Source set used for injecting the plugin classpath:
sourceSets.test
Any of these conventions can be reconfigured with the help of the class GradlePluginDevelopmentExtension.
The following Groovy-based sample demonstrates how to automatically inject the plugin classpath by using the standard conventions applied by the Java Gradle Plugin Development plugin.
plugins {
id 'groovy'
id 'java-gradle-plugin'
}
dependencies {
testImplementation('org.spockframework:spock-core:2.0-groovy-3.0') {
exclude group: 'org.codehaus.groovy'
}
}
plugins {
groovy
`java-gradle-plugin`
}
dependencies {
testImplementation("org.spockframework:spock-core:2.0-groovy-3.0") {
exclude(group = "org.codehaus.groovy")
}
}
Example: Automatically injecting the code under test classes into test builds
def "hello world task prints hello world"() {
given:
settingsFile << "rootProject.name = 'hello-world'"
buildFile << """
plugins {
id 'org.gradle.sample.helloworld'
}
"""
when:
def result = GradleRunner.create()
.withProjectDir(testProjectDir)
.withArguments('helloWorld')
.withPluginClasspath()
.build()
then:
result.output.contains('Hello world!')
result.task(":helloWorld").outcome == SUCCESS
}
The following build script demonstrates how to reconfigure the conventions provided by the Java Gradle Plugin Development plugin for a project that uses a custom Test
source set.
A new configuration DSL for modeling the below |
plugins {
id 'groovy'
id 'java-gradle-plugin'
}
def functionalTest = sourceSets.create('functionalTest')
def functionalTestTask = tasks.register('functionalTest', Test) {
group = 'verification'
testClassesDirs = sourceSets.functionalTest.output.classesDirs
classpath = sourceSets.functionalTest.runtimeClasspath
useJUnitPlatform()
}
tasks.named("check") {
dependsOn functionalTestTask
}
gradlePlugin {
testSourceSets sourceSets.functionalTest
}
dependencies {
functionalTestImplementation('org.spockframework:spock-core:2.0-groovy-3.0') {
exclude group: 'org.codehaus.groovy'
}
}
plugins {
groovy
`java-gradle-plugin`
}
val functionalTest = sourceSets.create("functionalTest")
val functionalTestTask = tasks.register<Test>("functionalTest") {
group = "verification"
testClassesDirs = functionalTest.output.classesDirs
classpath = functionalTest.runtimeClasspath
useJUnitPlatform()
}
tasks.check {
dependsOn(functionalTestTask)
}
gradlePlugin {
testSourceSets(functionalTest)
}
dependencies {
"functionalTestImplementation"("org.spockframework:spock-core:2.0-groovy-3.0") {
exclude(group = "org.codehaus.groovy")
}
}
Controlling the build environment
The runner executes the test builds in an isolated environment by specifying a dedicated "working directory" in a directory inside the JVM’s temp directory (i.e. the location specified by the java.io.tmpdir
system property, typically /tmp
). Any configuration in the default Gradle user home directory (e.g. ~/.gradle/gradle.properties
) is not used for test execution. The TestKit does not expose a mechanism for fine grained control of all aspects of the environment (e.g., JDK). Future versions of the TestKit will provide improved configuration options.
The TestKit uses dedicated daemon processes that are automatically shut down after test execution.
The dedicated working directory is not deleted by the runner after the build. The TestKit provides two ways to specify a location that is regularly cleaned, such as the project’s build folder:
-
The
org.gradle.testkit.dir
system property; -
The GradleRunner.withTestKitDir(file testKitDir) method.
The Gradle version used to test
The Gradle runner requires a Gradle distribution in order to execute the build. The TestKit does not depend on all of Gradle’s implementation.
By default, the runner will attempt to find a Gradle distribution based on where the GradleRunner
class was loaded from. That is, it is expected that the class was loaded from a Gradle distribution, as is the case when using the gradleTestKit()
dependency declaration.
When using the runner as part of tests being executed by Gradle (e.g. executing the test
task of a plugin project), the same distribution used to execute the tests will be used by the runner. When using the runner as part of tests being executed by an IDE, the same distribution of Gradle that was used when importing the project will be used. This means that the plugin will effectively be tested with the same version of Gradle that it is being built with.
Alternatively, a different and specific version of Gradle to use can be specified by the any of the following GradleRunner
methods:
This can potentially be used to test build logic across Gradle versions. The following demonstrates a cross-version compatibility test written as Groovy Spock test:
Example: Specifying a Gradle version for test execution
import org.gradle.testkit.runner.GradleRunner
import static org.gradle.testkit.runner.TaskOutcome.*
import spock.lang.TempDir
import spock.lang.Specification
import spock.lang.Unroll
class BuildLogicFunctionalTest extends Specification {
@TempDir File testProjectDir
File settingsFile
File buildFile
def setup() {
settingsFile = new File(testProjectDir, 'settings.gradle')
buildFile = new File(testProjectDir, 'build.gradle')
}
@Unroll
def "can execute hello world task with Gradle version #gradleVersion"() {
given:
buildFile << """
task helloWorld {
doLast {
logger.quiet 'Hello world!'
}
}
"""
settingsFile << ""
when:
def result = GradleRunner.create()
.withGradleVersion(gradleVersion)
.withProjectDir(testProjectDir)
.withArguments('helloWorld')
.build()
then:
result.output.contains('Hello world!')
result.task(":helloWorld").outcome == SUCCESS
where:
gradleVersion << ['5.0', '6.0.1']
}
}
Feature support when testing with different Gradle versions
It is possible to use the GradleRunner to execute builds with Gradle 1.0 and later. However, some runner features are not supported on earlier versions. In such cases, the runner will throw an exception when attempting to use the feature.
The following table lists the features that are sensitive to the Gradle version being used.
Feature | Minimum Version | Description |
---|---|---|
Inspecting executed tasks |
2.5 |
Inspecting the executed tasks, using BuildResult.getTasks() and similar methods. |
2.8 |
Injecting the code under test viaGradleRunner.withPluginClasspath(java.lang.Iterable). |
|
2.9 |
Inspecting the build’s text output when run in debug mode, using BuildResult.getOutput(). |
|
2.13 |
Injecting the code under test automatically via GradleRunner.withPluginClasspath() by applying the Java Gradle Plugin Development plugin. |
|
Setting environment variables to be used by the build. |
3.5 |
The Gradle Tooling API only supports setting environment variables in later versions. |
Debugging build logic
The runner uses the Tooling API to execute builds. An implication of this is that the builds are executed in a separate process (i.e. not the same process executing the tests). Therefore, executing your tests in debug mode does not allow you to debug your build logic as you may expect. Any breakpoints set in your IDE will be not be tripped by the code being exercised by the test build.
The TestKit provides two different ways to enable the debug mode:
-
Setting “
org.gradle.testkit.debug
” system property totrue
for the JVM using theGradleRunner
(i.e. not the build being executed with the runner); -
Calling the GradleRunner.withDebug(boolean) method.
The system property approach can be used when it is desirable to enable debugging support without making an adhoc change to the runner configuration. Most IDEs offer the capability to set JVM system properties for test execution, and such a feature can be used to set this system property.
Testing with the Build Cache
To enable the Build Cache in your tests, you can pass the --build-cache
argument to GradleRunner or use one of the other methods described in Enable the build cache. You can then check for the task outcome TaskOutcome.FROM_CACHE when your plugin’s custom task is cached. This outcome is only valid for Gradle 3.5 and newer.
Example: Testing cacheable tasks
def "cacheableTask is loaded from cache"() {
given:
buildFile << """
plugins {
id 'org.gradle.sample.helloworld'
}
"""
when:
def result = runner()
.withArguments( '--build-cache', 'cacheableTask')
.build()
then:
result.task(":cacheableTask").outcome == SUCCESS
when:
new File(testProjectDir, 'build').deleteDir()
result = runner()
.withArguments( '--build-cache', 'cacheableTask')
.build()
then:
result.task(":cacheableTask").outcome == FROM_CACHE
}
Note that TestKit re-uses a Gradle user home between tests (see GradleRunner.withTestKitDir(java.io.File)) which contains the default location for the local build cache. For testing with the build cache, the build cache directory should be cleaned between tests. The easiest way to accomplish this is to configure the local build cache to use a temporary directory.
Example: Clean build cache between tests
@TempDir File testProjectDir
File buildFile
File localBuildCacheDirectory
def setup() {
localBuildCacheDirectory = new File(testProjectDir, 'local-cache')
buildFile = new File(testProjectDir,'settings.gradle') << """
buildCache {
local {
directory '${localBuildCacheDirectory.toURI()}'
}
}
"""
buildFile = new File(testProjectDir,'build.gradle')
}