Support free tutorials





Quick links

Developing Android unit and instrumentation tests - Tutorial

Lars Vogel

Version 2.0

09.09.2015

Android Testing

This tutorial describes how to write unit and instrumentation tests for your Android application. It also cover the usage of Mockito for Android tests and describes how to execute these tests via Android studio and Gradle. Links are provides for other test frameworks like Espresso.


Table of Contents

1. Why is testing important for Android applications?
1.1. Why is testing especially important for Android application?
1.2. Android test strategy
1.3. Android and testing support
2. Prerequisites
3. Android automated testing
3.1. What to test on Android applications
3.2. Testing preconditions
3.3. Testing a single app or multiple apps
4. Android unit and instrumentation unit tests
4.1. Categories of Android unit tests
4.2. Unit tests (local tests)
4.3. Instrumented tests for testing Java classes which use the Android API
5. Android project structure and test folder creation
5.1. Android project organization for tests
6. Running tests and solving issues with running tests
6.1. Selecting which tests are compiled
6.2. Seeing all passed tests your test builds
6.3. Build error in your test builds
7. Unit testing on the JVM
7.1. Unit testing in Android
7.2. Location of unit tests
7.3. Required dependencies in the Gradle build file
7.4. Run the unit tests from Gradle
7.5. Run the unit tests from Android Studio
7.6. Location of test reports
7.7. Activating default return values for mocked methods in android.jar
8. Exercise preparation: Create Android project to test project
9. Exercise: Create unit test for your Android project
9.1. Target of this exercise
9.2. Add JUnit dependency
9.3. Create test
9.4. Run unit tests
10. Instrumentation - The underlying Android testing API
10.1. Instrumentation
10.2. How the Android system executes tests
10.3. Usage of the instrumentation framework
11. Instrumented unit testing
11.1. Using instrumented unit tests in Android
11.2. Location of instrumentation tests
11.3. Define dependencies and testInstrumentationRunner in the Gradle build file
11.4. Using the @RunWith(AndroidJUnit4.class)
11.5. Run the unit tests from Gradle
11.6. Run the unit tests from Android Studio
11.7. Location of test reports
12. Exercise: Create simple instrumented unit test
13. Mocking objects in Android
14. Using Mockito
15. Using Mockito on Android
16. Exercise: Mocking file access
16.1. Create class to test
16.2. Create a new unit test
17. Android Testing Support Library
18. More on Android testing
18.1. Android additional assertion
18.2. Test groups
18.3. Test filtering
18.4. Flaky tests
19. Activity testing
20. Using the Monkey tool for creating a random event stream
20.1. What is monkey?
20.2. How to use Monkey
21. Other Open Source testing frameworks
22. Application testing
23. Exercise: Testing the Android application
23.1. Create project
23.2. Create unit test for application object
23.3. Create unit test for the application object
23.4. Create instrumented test for application object
24. Service testing
25. Content provider testing
26. Loader testing
27. Test folder creation in Android Studio
28. About this website
29. Links and Literature
29.1. Android testing resources
29.2. vogella GmbH training and consulting support

1. Why is testing important for Android applications?

1.1. Why is testing especially important for Android application?

Android applications run on devices with limited memory, CPU power and power supply. The behavior of the application also depends on external factors like connectivity, general system utilization, etc.

Therefore, it is very important to debug, test and optimize your Android application. Having a reasonable test coverage for your Android application helps you to enhance and maintain the Android application.

1.2. Android test strategy

As it is not possible to test Android applications on all possible device configurations, it is a common practice to run Android tests on typical device configurations. You should test your application at least on one device with the lowest possible configuration and on one device with the highest available configuration, e.g., pixel density, screen resolution to ensure that it works fine on these devices.

1.3. Android and testing support

In 2015 the tool and framework support for Android applications was hugely improved. The Android test system has been updated to be based on JUnit 4 and you can run unit tests either on the Java virtual machine or on an Android runtime. Also, Google introduced a user interface testing framework called Espresso which makes is easy to develop user interface tests for Android applications.

2. Prerequisites

The following assumes that you have already a basic understanding of how Android applications are created. See the Android Tutorial for details. Also the following description assume that you know the basis of the Android build system based on Gradle. See Building Android applications with Gradle Tutorial for an introduction.

For more information about Gradle see the Gradle Tutorial.

3. Android automated testing

3.1. What to test on Android applications

In general you should focus your Android tests on testing the business logic of your application. A good rule of thumb is to have the following distribution of tests:

  • 70-80 % unit tests to ensure stability of your code basis

  • 20-30 % functional tests to ensure that the application really works

  • some cross functional tests if your application integrates intensively with other Application components

3.2. Testing preconditions

It is good practice in Android testing to have one method called testPreconditions() which tests the pre-conditions for all other tests. If this method fails, you know immediately that the assumptions for the other tests have been violated.

3.3. Testing a single app or multiple apps

Another important criteria for testing is if you only test your own application or if you test the integration with other applications. If you run tests within your application you can use testing frameworks which require some knowledge about your application, e.g., the view IDs.

4. Android unit and instrumentation unit tests

4.1. Categories of Android unit tests

Android testing is based on JUnit. Unit testing for Android can be classified into:

  • Local unit tests - tests which can run on the JVM

  • Instrumented unit tests - tests which require the Android system

Android testing categories

If possible, you should prefer to use local tests as the test execution is much faster compared to the time required to deploy and run the test on an Android device.

4.2. Unit tests (local tests)

The Android Gradle plug-in supports running Android unit tests on the JVM. To enable this, this Gradle plug-in creates a special version of the android.jar (also known as the Android mockable jar) and provides it to the unit test so that all fields, methods and classes are available. Any call to the Android mockable JAR result, by default, in an exception.

Therefore, if your classes do not call the Android API or have only minimal dependencies to it, you can use the JUnit test framework (or any other Java testing framework) without any restrictions. Any Android dependency in your code must be replaced for the unit tests, e.g., via a mocking framework like Mockito. Also see Section 7.7, “Activating default return values for mocked methods in android.jar” how to adjust that to get certain defaults returned from the Android mockable jar.

The advantages of running your tests on the JVM is that the execution speed of the tests is extremely fast compared to tests running on Android.

4.3. Instrumented tests for testing Java classes which use the Android API

If you want to test code which use the Android API, you need to run these tests on an Android device. Unfortunately, this makes the execution of tests take longer.

This is because android.jar JAR file mocked by the Android tooling does not execute the real Android code. In its default configuration it throws exceptions.

5. Android project structure and test folder creation

5.1. Android project organization for tests

The preferred way of organizing tests is based on a convention. In your application project you should use the following base folder structure for your code organization, this is also the structure the project wizard creates.

  • app/src/main/java - for your source code of your main application build

  • app/src/test/java - for any unit test which can run on the JVM

  • app/src/androidTest/java - for any test which should run on an Android device

If you follow this conversion than the Android build system can automatically run the unit tests on the JVM and the Android tests on the Android device.

6. Running tests and solving issues with running tests

6.1. Selecting which tests are compiled

Warning

To make Android Studio aware of your new dependency for unit or instrumentation tests you may have to select Unit Tests or Android Instrumentation Tests from the Build Variant view in Android Studio.

Making Android Studio aware of new dependencies

How you develop your tests depends if you are developing unit tests or integration tests.

6.2. Seeing all passed tests your test builds

To see all passed tests, click on the highlighted toolbar entry in the following screenshot.

Android Studio see all tests

6.3. Build error in your test builds

If you receive the following error message: "error duplicate files in path. Path in archive: LICENSE.txt" you can add the following to your app/gradle.build file.

android {
    packagingOptions {
    exclude 'LICENSE.txt'
    }
} 

7. Unit testing on the JVM

7.1. Unit testing in Android

Android uses the term unit tests for tests which can run on a local JVM on the development machine instead of the Android Runtime.

A unit test tests only the functionality of a certain component. For example, assume a button in an Android activity is used to start another activity. A unit test would determine if the corresponding intent was issued, not if the second activity was started.

The unit tests are executed against a modified version of the android.jar Android library where all final modifiers have been stripped off. This allow you to mocking libraries, like Mockito. All methods in the used android.jar file throw exceptions (by default). This is to make sure your unit tests only test your code and do not depend on any particular behavior of the Android platform. If you want specific behavior you can use a mocking framework to replace these call.

7.2. Location of unit tests

As described in androidtesting_projectstructure the unit tests of an Android project should be located in the app/src/test folder. See Section 27, “Test folder creation in Android Studio” for the creation of such a test folder.

7.3. Required dependencies in the Gradle build file

To use JUnit tests for your Android application you need to add the dependency to JUnit and you other test dependencies to your Gradle build file.

dependencies {
    // Unit testing dependencies
    testCompile 'junit:junit:4.12'
    // Set this dependency if you want to use the Hamcrest matcher library
    testCompile 'org.hamcrest:hamcrest-library:1.3'
    // more stuff, e.g., Mockito
} 

7.4. Run the unit tests from Gradle

Run your unit tests wit the gradlew test command.

7.5. Run the unit tests from Android Studio

Android Studio has two types of artifacts which can be selected in the build variant view. If you select Unit Tests in this view, the unit tests are executed on the JVM.

Running the Android tests from Android Studio

To run the unit, ensure that you have select Unit tests, right-click on your test class in the Project window and select Run.

Running Unit tests in Android Studio

7.6. Location of test reports

The rest reports are created in the app/build/reports/tests/debug/ directory. The index.html gives an overview and links to the individual test pages.

7.7. Activating default return values for mocked methods in android.jar

You can also instruct the Gradle build system to return default values for method calls in the android.jar with the following configuration in your Gradle build file.

android {
  // ...
  testOptions { 
    unitTests.returnDefaultValues = true
  }
} 

8. Exercise preparation: Create Android project to test project

Create the Android project as described in Android temperature converter.

9. Exercise: Create unit test for your Android project

9.1. Target of this exercise

In this exercise you learn how to create a JUnit4 test for the Android project you created in Section 8, “Exercise preparation: Create Android project to test project”.

9.2. Add JUnit dependency

Ensure you have the dependency to Junit in your app/build.gradle file. If the test folder structure is missing in your project, follow the process described in Section 27, “Test folder creation in Android Studio” to create it.

dependencies {
  // Unit testing dependencies
  testCompile 'junit:junit:4.12'
} 

9.3. Create test

In your app/src/test directory create the following two test methods of the ConverterUtil class.

package com.vogella.android.temperature.test;

import static org.junit.Assert.*;

import org.junit.After;
import org.junit.Before;
import org.junit.Test;

import com.vogella.android.temperature.ConverterUtil;

public class ConverterUtilTest {

  @Test
  public void testConvertFahrenheitToCelsius() {
    float actual = ConverterUtil.convertCelsiusToFahrenheit(100);
    // expected value is 212
    float expected = 212;
    // use this method because float is not precise
    assertEquals("Conversion from celsius to fahrenheit failed", expected,
        actual, 0.001);
  }

  @Test
  public void testConvertCelsiusToFahrenheit() {
    float actual = ConverterUtil.convertFahrenheitToCelsius(212);
    // expected value is 100
    float expected = 100;
    // use this method because float is not precise
    assertEquals("Conversion from celsius to fahrenheit failed", expected,
        actual, 0.001);
  }

} 

9.4. Run unit tests

If everything is implemented correctly the tests should execute successfully.

10. Instrumentation - The underlying Android testing API

10.1. Instrumentation

The Android testing API provides hooks into the Android component and application life cycle. These hooks are called the instrumentation API and allow your tests to control the life cycle and user interaction events.

Under normal circumstances your application can only react to the life cycle and user interaction events. For example if Android creates your activity the onCreate() method is called on your activity. Or the user presses a button or a key and your corresponding code is called. Via instrumentation you can control these events via your tests.

Only an instrumentation-based test class allows you to send key events (or touch events) to the application under test.

For example, your test can call the getActivity() method which starts an activity and returns the activity under test. Afterwards, you can call the finish() method, followed by a getActivity() method call again and you can test if the application restored its state correctly.

10.2. How the Android system executes tests

The InstrumentationTestRunner is the base test runner for Android tests. This test runner starts and loads the test methods. Via the instrumentation API it communicates with the Android system. If you start a test for an Android application, the Android system kills any process of the application under test and then loads a new instance. It does not start the application, this is the responsibility of the test methods. The test method controls the life cycle of the components of the application.

The test runner also calls the onCreate() method of the application and activity under test during its initialization.

10.3. Usage of the instrumentation framework

With unit testing for Android which run their unit tests directly on the JVM and popular user interface testing framework like Espresso the developer rarely has to use the instrumentation API directly.

11. Instrumented unit testing

11.1. Using instrumented unit tests in Android

Instrumented unit tests are unit tests that run on Android devices and emulators instead of running on the Java virtual machine. These tests have access to the real device and its resources and are useful to unit test functionality which cannot be easily mocked by mocking frameworks. Also developer uses these tests if the framework functionality must be used to validate a functionality. Example are for example a test which validates a Parelable implementation.

Mockito as mocking framework can still be used to mock the parts of the Android system which are not interesting for the test.

11.2. Location of instrumentation tests

As described in androidtesting_projectstructure the unit tests of an Android project should be located in the app/src/androidTest/java folder. See Section 27, “Test folder creation in Android Studio” for the creation of such a test folder.

11.3. Define dependencies and testInstrumentationRunner in the Gradle build file

To use JUnit tests for your Android application you need to add the dependency to JUnit and you other test dependencies to your Gradle build file. You needs also to specify the default testInstrumentationRunner runner as "android.support.test.runner.AndroidJUnitRunner".

defaultConfig {
       ..... more stuff
        testInstrumentationRunner "android.support.test.runner.AndroidJUnitRunner"
    }

dependencies {
    // Unit testing dependencies
    androidTestCompile 'junit:junit:4.12'
    // Set this dependency if you want to use the Hamcrest matcher library
    androidTestCompile 'org.hamcrest:hamcrest-library:1.3'
    // more stuff, e.g., Mockito
} 

11.4. Using the @RunWith(AndroidJUnit4.class)

It is also recommended annotating the test with the @RunWith(AndroidJUnit4.class) annotation. AndroidJUnit4 extends JUnit4, so if you use pure Junit4 syntax and ActivityTestRule it is not required. But you need it if you want to run, i.e., Espresso tests with ActivityTestRule and JUnit4.

11.5. Run the unit tests from Gradle

Run your unit tests wit the gradlew connectedCheck command.

11.6. Run the unit tests from Android Studio

Android Studio has two types of artifacts which can be selected in the build variant window. If you select Android Instrumentation Tests the unit tests are executed on the JVM.

Running the Android instrumentation tests from Android Studio

To run the unit, ensure that you have select Android Instrumentation Tests, right-click on your test class in the Project window and select Run.

11.7. Location of test reports

The rest reports are created in the app/build/reports/androidTests/connected/ directory. The index.html gives an overview and links to the individual test pages.

12. Exercise: Create simple instrumented unit test

You will later learn to create more complex tests but for now duplicate your test you created in Section 9, “Exercise: Create unit test for your Android project” and also run it as instrumentation test.

13. Mocking objects in Android

For you unit tests you can use mocking a framework to mock Android objects. This works also for tests running the a Android runtime. The Android framework provided in the past specialized mocking classes but these are not necessary anymore.

14. Using Mockito

See Mockito tutorial for information how to use the Mockito framework.

15. Using Mockito on Android

Mockito can also be directly used in Android unit tests simply by adding the dependency to Mockito to the Gradle build file of the application. To use it in instrumented Android tests (since the release 1.9.5). Which requires that dexmaker and dexmaker-mockito are also added as dependency. in the Gradle build file.

dependencies {
    testCompile 'junit:junit:4.12'
    // required if you want to use Mockito for unit tests
    testCompile 'org.mockito:mockito-core:1.+'
    // required if you want to use Mockito for Android instrumentation tests
    androidTestCompile 'org.mockito:mockito-core:1.+'
    androidTestCompile "com.google.dexmaker:dexmaker:1.2"
    androidTestCompile "com.google.dexmaker:dexmaker-mockito:1.2"
} 

16. Exercise: Mocking file access

16.1. Create class to test

In an existing or in a new Android application with the package com.vogella.android.testing.mockitocontextmock add the following class. It is not required to use this class in your Android application, as it has only be provided to demonstrate the usage of Mockito for a unit test.

public class Util {
  public static void writeConfiguration(Context ctx) {
    BufferedWriter writer = null;
    try {
      FileOutputStream openFileOutput = 
       ctx.openFileOutput("config.txt", Context.MODE_PRIVATE);
      openFileOutput.write("This is a test1.".getBytes());
      openFileOutput.write("This is a test2.".getBytes());
    } catch (Exception e) {
      throw new RuntimeException(e);
    } finally {
      if (writer != null) {
        try {
          writer.close();
        } catch (IOException e) {
          e.printStackTrace();
        }

      }
    }
  }
} 

Note

16.2. Create a new unit test

Using Mockito the mocking framework, write a unit test which validates that:

  • openFileOutput is called exactly once

  • the write() method is called at least twice.

package com.vogella.android.testing.mockitocontextmock;

import android.content.Context;

import org.junit.Before;
import org.junit.Test;
import org.junit.runner.RunWith;
import org.mockito.Mock;
import org.mockito.MockitoAnnotations;

import java.io.FileOutputStream;

import static org.junit.Assert.fail;
import static org.mockito.Matchers.any;
import static org.mockito.Matchers.anyInt;
import static org.mockito.Matchers.anyString;
import static org.mockito.Mockito.atLeast;
import static org.mockito.Mockito.times;
import static org.mockito.Mockito.verify;
import static org.mockito.Mockito.when;

public class TextContextOutputStream {

@Mock
Context context;

@Mock
FileOutputStream fileOutputStream;

@Before
public void init(){
    MockitoAnnotations.initMocks(this);
}

    @Test
public void writeShouldWriteTwiceToFileSystem() {
    try {
        when(context.openFileOutput(anyString(), anyInt())).thenReturn(fileOutputStream);
        Util.writeConfiguration(context);
        verify(context, times(1)).openFileOutput(anyString(), anyInt());
        verify(fileOutputStream, atLeast(2)).write(any(byte[].class));

    } catch (Exception e) {
        e.printStackTrace();
        fail();
    }
}
} 

17. Android Testing Support Library

The Android Testing Support Library provides functionality to create and run Android tests. The library contains the AndroidJUnitRunner, the Espresso test framework and the UI Automator.

AndroidJUnitRunner allows to create and run JUnit 4 tests, while the Espresso test framework can be used to test the User Interface of your application. UI Automator allows to write cross application functional tests.

AndroidJunitRunner provides access to the instrumentation API, via the InstrumentationRegistery.

  • InstrumentationRegistry.getInstrumentation(), returns the Instrumentation currently running.

  • InstrumentationRegistry.getContext(), returns the Context of this Instrumentation’s package.

  • InstrumentationRegistry.getTargetContext(), returns the application Context of the target application.

  • InstrumentationRegistry.getArguments(), returns a copy of arguments Bundle that was passed to this Instrumentation. This is useful when you want to access the command line arguments passed to Instrumentation for your test.

It also gives access to the life cycle via the ActivityLifecycleMonitorRegistry.

18. More on Android testing

18.1. Android additional assertion

The Android testing API provides the MoreAsserts and ViewAsserts classes in addition to the standard JUnit Assert class.

18.2. Test groups

The @SmallTest, @MediumTest and @LargeTest annotations allows to classify tests.

This allows you to run, for example, only tests which do not run very long. Long running tests could than run only in the continuous integration server.

To run only selected tests you can configure the InstrumentationTestRunner via your Gradle plug-in. The following listing is an example for your build.gradle file to run only the tests annotated with @SmallTests.

android {
  //....
  defaultConfig {
  //....
    testInstrumentationRunner "android.support.test.runner.AndroidJUnitRunner"
    testInstrumentationRunnerArgument "size", "small"
  }
} 

Note

This works only as of the Android Plugin for Gradle 1.3.0. Also this is currently not yet supported in Android Studio, i.e., if you run a test from Android Studio the selection in the Gradle build file is ignored.

import android.test.suitebuilder.annotation.MediumTest;
import android.test.suitebuilder.annotation.SmallTest;

import org.junit.Test;

public class ExampleTest {

    @Test
    @SmallTest
    public void validateSecondActivity() {
        // Do something not so long...
    }

    @Test
    @MediumTest
    public void validateSecondActivityAgain() {
        // Do something which takes more time....
    }

} 

18.3. Test filtering

You can annotate tests with annotations. The Android test runner allows to filter these tests.

Table 1. Annotations for filtering tests

Annotation Description
@RequiresDevice: Specifies that the test should run only on physical devices, not on emulators.
@SdkSupress: @SDKSupress(minSdkVersion=18)


18.4. Flaky tests

Actions in Android are sometimes time dependent. To tell Android to repeat a test once it fails, use the @FlakyTest annotation. Via the tolerance attribute of this annotation you can define how often the Android test framework should try to repeat a test before marking it as failed.

19. Activity testing

See Android user interface testing with Espresso and Cross component testing with UI Automator.

20. Using the Monkey tool for creating a random event stream

20.1. What is monkey?

Monkey is a command line tool which sends pseudo random events to your device. You can restrict Monkey to run only for a certain package and therefore instruct Monkey to test only your application.

20.2. How to use Monkey

For example, the following will send 2000 random events to the application with the de.vogella.android.test.target package.

adb shell monkey -p de.vogella.android.test.target -v 2000 

Monkey sometimes causes problems with the adb server. Use the following commands to restart the adb server.

adb kill-server
adb start-server 

You can use the -s [seed] parameter to ensure that the generated sequence of events is always the same.

For more info on Monkey please see the Monkey description.

21. Other Open Source testing frameworks

Robotium is an Open Source framework on top of the Android testing framework which makes the testing API simpler. See Robotium for user interface testing with the Robotium framework.

Robolectric is an Open Source framework which allows you to run tests which use the Android API directly on the JVM. See the Robolectric tutorial for more information.

Roboguice allows you to use dependency injection in your Android components which simplifies testing. See Using Roboguice.

22. Application testing

The application class contains the logic, data and settings which are relevant for the whole application. Therefore you should test this object, to ensure it works correctly.

You can write JUnit 4 for the application object and test it on the JVM. In this case you would mock all dependencies to the application object.

To test an Android application object on the Android runtime you use the ApplicationTestCase class. It is expected that Google will soon provide a special JUnit4 rule for testing the application object but at the moment his is not yet available.

The test runner of the Android tests (InstrumentationTestRunner) creates automatically an instance of application during its initialization phase. If you do asynchronous processing in your onCreate method you should consider that.

23. Exercise: Testing the Android application

23.1. Create project

Create a new Android application with the com.vogella.android.testing.applicationtest package name based on the Blank Activity template.

Add the following application object to your application.

package com.vogella.android.testing.applicationtest;

import android.app.Application;

import java.util.ArrayList;
import java.util.List;

public class MyApplication extends Application {
    public static final List<String> list = new ArrayList<String>();
} 

Also declare the application it in your manifest file.

<?xml version="1.0" encoding="utf-8"?>
<manifest xmlns:android="http://schemas.android.com/apk/res/android"
    package="com.vogella.android.testing.applicationtest" >

    <application
        android:name=".MyApplication"
        android:allowBackup="true"
        android:icon="@mipmap/ic_launcher"
        android:label="@string/app_name"
        android:theme="@style/AppTheme" >
        <activity
            android:name=".MainActivity"
            android:label="@string/app_name" >
            <intent-filter>
                <action android:name="android.intent.action.MAIN" />

                <category android:name="android.intent.category.LAUNCHER" />
            </intent-filter>
        </activity>
    </application>

</manifest> 

23.2. Create unit test for application object

In your app/src/test directory a new unit test and assert that the MyApplication.list field is not empty and has initially a size of zero.

23.3. Create unit test for the application object

Create the following unit test for your application object.

package com.vogella.android.testing.applicationtest;

import android.content.pm.PackageInfo;
import android.test.ApplicationTestCase;
import android.test.MoreAsserts;

public class ApplicationTest extends ApplicationTestCase<MyApplication> {

    private MyApplication application;

    public ApplicationTest() {
        super(MyApplication.class);
    }

    protected void setUp() throws Exception {
        super.setUp();
        createApplication();
        application = getApplication();

    }

    public void testCorrectVersion() throws Exception {
        PackageInfo info = application.getPackageManager().getPackageInfo(application.getPackageName(), 0);
        assertNotNull(info);
        MoreAsserts.assertMatchesRegex("\\d\\.\\d", info.versionName);
    }

} 

23.4. Create instrumented test for application object

Create the following unit test based on the JUnit 3 testing framework.

package com.vogella.android.testing.applicationtest;

import android.content.pm.PackageInfo;
import android.test.ApplicationTestCase;
import android.test.MoreAsserts;

public class ApplicationTest extends ApplicationTestCase<MyApplication> {

    private MyApplication application;

    public ApplicationTest() {
        super(MyApplication.class);
    }

    protected void setUp() throws Exception {
        super.setUp();
        createApplication();
        application = getApplication();

    }

    public void testCorrectVersion() throws Exception {
        PackageInfo info = application.getPackageManager().getPackageInfo(application.getPackageName(), 0);
        assertNotNull(info);
        MoreAsserts.assertMatchesRegex("\\d\\.\\d", info.versionName);
    }

} 

24. Service testing

To test a service you use the ServiceTestRule class from the Android Testing Support Library. The old ServiceTestCase is deprecated.

This rule provides a simplified mechanism to start and shutdown your service before and after your test. It guarantees that the service is successfully connected when starting (or binding to) a service. The service can be started (or bound) using one of the helper methods. It will automatically be stopped (or unbound) after the test completes and any methods annotated with @After are finished.

Note

Note: This rule doesn't support IntentService because it's automatically destroyed after the onHandleIntent method.

The following listing is an example of testing a service.

@RunWith(AndroidJUnit4.class)
@MediumTest
public class MyServiceTest {

    @Rule
    public final ServiceTestRule mServiceRule = new ServiceTestRule();

    // test for a service which is started with startService
    @Test
    public void testWithStartedService() {
        mServiceRule.
        startService(new Intent(InstrumentationRegistry.getTargetContext(), 
            MyService.class));
        // test code
    }

    @Test
 // test for a service which is started with bindService
    public void testWithBoundService() {
        IBinder binder = mServiceRule.
            bindService(new Intent(InstrumentationRegistry.getTargetContext(), 
                MyService.class));
        MyService service = ((MyService.LocalBinder) binder).getService();
        assertTrue("True wasn't returned", service.doSomethingToReturnTrue());
    }
} 

25. Content provider testing

To test a content provider, you use the ProviderTestCase2 class. ProviderTestCase2 automatically instantiates the provider under test and inserts an IsolatedContext object which is isolated from the Android system, but still allows file and database access.

The usage of the IsolatedContext object ensures that your provider test does not affect the real device.

ProviderTestCase2 also provides access to a MockContentResolver via the getMockContentResolver() method.

You should test all operations of the provider and also what happens if the provider is called with an invalid URI or with an invalid projection.

26. Loader testing

To test a loader, you use the LoaderTestCase class. It is expected that a JUnit 4 rule will be provided in the future to replace this class.

27. Test folder creation in Android Studio

The most recent version of Android Studio has added a test folder to its default project template. In case you using a template which does not create a test folder, you have to create it manually. To create the test folder in Android Studio, switch to the Project view, this shows you the directory structure of your project.

Switching to the Project view in Android Studio

Select the src and use the context menu to create a new test folder.

Switching to the Project view in Android Studio

Adding a java folder

Adding a java folder

If not yet done, also add the JUnit dependency to your Gradle build file.

dependencies {
    // Unit testing dependencies
    testCompile 'junit:junit:4.12'
    // Set this dependency if you want to use the Hamcrest matcher library
    testCompile 'org.hamcrest:hamcrest-library:1.3'
    // more stuff, e.g., Mockito
} 

Warning

The creation of the Java folder might add the new test directory as source file to your build.gradle file. If you have the following entry in your app/build.gradle file you need to REMOVE it. test should NOT be treated as normale source folder.

sourceSets { main { java.srcDirs = ['src/main/java', 'src/test/java/'] } } 

Afterwards you can add your unit test to this folder structure.

28. About this website

29. Links and Literature

29.2. vogella GmbH training and consulting support

TRAINING SERVICE & SUPPORT
The vogella company provides comprehensive training and education services from experts in the areas of Eclipse RCP, Android, Git, Java, Gradle and Spring. We offer both public and inhouse training. Whichever course you decide to take, you are guaranteed to experience what many before you refer to as “The best IT class I have ever attended”. The vogella company offers expert consulting services, development support and coaching. Our customers range from Fortune 100 corporations to individual developers.