JUnit

AIO Tests allows easy reporting of JUnit results by supporting the import of the TEST-feature-results.xml (the output file of a JUnit execution).

With AIO Tests JUnit integration, developers will not need to write manual cases inside AIO Tests. They can write JUnit cases which will be imported (created) in AIO Tests along with their execution results. This results in a holistic view of the coverage of requirements (unit to manual to functional and end-to-end automated tests).

This article discusses how to import JUnit XML results and map them to existing cases in AIO Tests or generate new cases in AIO, in the absence of existing cases.

In this documentation, you’ll understand:

JUnit

JUnit is one of the most widely used open-source Unit Testing frameworks for JAVA and has been a key player in promoting test-driven development.

The latest version of JUnit, JUnit 5 has introduced many new features, new APIs, and the concept of a test engine to simplify test automation (e.g. @DisplayName, @BeforeXX, @ParametrizedTest, etc.).

JUnit Reporting

JUnit’s XML report comes through the Maven Surefire plugin or through an ANT task.

Being one of the earliest testing frameworks and with a huge community base, JUnit’s XML report has gained widespread adoption. Many continuous integration servers accept JUnit XML reports as their de facto standard for reporting test results.

The report format gradually became ubiquitous for test results reporting and frameworks across languages like PyTest, Cucumber, Fitnesse, RSpec, Katalon, etc. All these have reporting extensions that generate the JUnit XML report, which implies cases from any of these frameworks can be imported into AIO Tests.

Sample JUnit XML Report

<?xml version="1.0" encoding="UTF-8"?>
<testsuite xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" 
xsi:noNamespaceSchemaLocation="https://maven.apache.org/surefire/maven-surefire-plugin/xsd/surefire-test-report-3.0.xsd" version="3.0" name="com.aio.tests.junit5.ForestCreatorTests" time="0.003" tests="2" errors="0" skipped="0" failures="1">
  <properties>
    <property name="gopherProxySet" value="false"/>
    ....
  </properties>
  <testcase name="AT-TC-26, AT-TC-27,AT-TC-28 : Verify positive case - all trees get planted" classname="com.aio.tests.junit5.ForestCreatorTests" time="0.001">
    <failure type="org.opentest4j.AssertionFailedError"><![CDATA[org.opentest4j.AssertionFailedError: Verify end to end plant growth ==> expected: <10> but was: <1>
   at com.aio.tests.junit5.ForestCreatorTests.testCreateForest(ForestCreatorTests.java:24)
]]></failure>
    <system-out><![CDATA[Planting tree 0
Watering the plants 
Removing the weeds 
]]></system-out>
  </testcase>
  <testcase name="SCRUM-TC-25 : Verify positive case - all trees 
  get planted" classname="com.aio.tests.junit5.ForestCreatorTests" time="0.001">
    <system-out><![CDATA[Planting tree 0
...
]]></system-out>
  </testcase>
</testsuite>

Generating JUnit XML Report

Maven Surefire

The Maven surefire plugin is required to generate the JUnit XML Report.

JUnit 4 Settings

<build>
    <plugins>
        <plugin>
            <groupId>org.apache.maven.plugins</groupId>
            <artifactId>maven-surefire-plugin</artifactId>
            <version>3.0.0-M5</version>
        </plugin>
        ...
    </plugins>
</build>

JUnit 5 Settings

If the JUnit 5 features are being utilized, the junit-jupiter-engine needs to be added with additional configurations for fine-grained configuration of reports.
Tests need to be annotated with @DisplayName and <usePhrasedXXXName>true</usePhrasedXXXName> fields need to be set to true as shown below.

For more info, refer to Maven Surefire - Using JUnit 5

<plugin>
    <groupId>org.apache.maven.plugins</groupId>
    <artifactId>maven-surefire-plugin</artifactId>
    <version>3.0.0-M5</version>
    <dependencies>
        <dependency>
            <groupId>org.junit.jupiter</groupId>
            <artifactId>junit-jupiter-engine</artifactId>
            <version>5.7.0-M1</version>
        </dependency>
    </dependencies>
    <configuration>
                <statelessTestsetReporter implementation="org.apache.maven.plugin.surefire.extensions.junit5.JUnit5Xml30StatelessReporter">
                    <disable>false</disable>
                    <version>3.0</version>
                    <usePhrasedFileName>true</usePhrasedFileName>
                    <usePhrasedTestSuiteClassName>true</usePhrasedTestSuiteClassName>
                    <usePhrasedTestCaseClassName>true</usePhrasedTestCaseClassName>
                    <usePhrasedTestCaseMethodName>true</usePhrasedTestCaseMethodName>
                </statelessTestsetReporter>
.....
      </configuration>
 </plugin>

Ant

If you are using Ant, the following tasks need to be used to generate the report.

JUnit 4 - JUnit Task

JUnit 5 - JUnitlauncher Task

Using the junitlauncher task allows launching the JUnit 5 test launcher and building the test requests to be executed by the test engine(s) supported by JUnit 5. For more information, please refer to JUnitLauncher.

XML generated by this formatter can be used as-is by the junitreport task.

Status Mapping JUnit → AIO Tests

JUnit XML	Description	AIO Tests Mapping

JUnit XML	Description	AIO Tests Mapping
No tag inside <testcase> means Passed	Passed case	Passed
</skipped>	Skipped case either by @Ignore or others	Not Run
</failure>	`Indicates that the test failed. A failure is a test which the code has explicitly failed by using the mechanisms for that purpose. e.g., via an assertEquals.`	Failed
</error>	`Indicates that the test errored. An errored test is one that had an unanticipated problem. e.g., an unchecked throwable;`	Failed

Mapping automated JUnit tests to AIO Tests

In going with the simplicity advantage of AIO Tests, the JUnit integration has been designed as a simple and non-intrusive integration, without any extra dependency or coding required.

There are 4 ways that a JUnit case is mapped to an AIO Case.

JUnit 5: @DisplayName gives a way to specify generic information in the test name instead of the hard rule-based Java method names. This can be used to specify an existing AIO Case Key.
For example, @DisplayName("SCRUM-TC-1973 : Test forest gets created.")
Testcase Key in Underscores: Since Java and Python, do not allow hyphens in their method signatures, AIO Tests allows case keys with underscores as valid mapping. The AIO Test key can be used in the method name by replacing the hyphen with underscores. eg. if AT-TC-123 is being automated by a test, the test method can be named verifyNotNullType_AT_TC_123 and the results of this method will be marked against AT-TC-123.
Automation Key: The mapping of an AIO Case to an automated JUnit case can happen through a field Automation Key, which can be specified in the Case in AIO Tests app as shown below. The automation key is the fully qualified name of the test method i.e. packagename.classname.methodname.

This can be useful at places where you do not want to add AIO case keys in the name or annotations or where automation is happening first without a manual case being created in AIO Tests.

New Case Creation: If no case key (e.g. SCRUM-TC-xx) or automation key is found with existing cases, a new case is created with:
- Title as the name value from <testcase> tag of the JUnit report
- Automation key as classname.name from the JUnit report.
- Status as Published
- Automation status as Automated
- Automation owner as a user uploading the results

The following flow summarizes the logic of JUnit case mapping when a key is not found via @DisplayName and via the Automation key.

Examples

Below are a few examples that show the results of testing the ForestCreator.java class.

Mapping using JUnit5 @DisplayName (when settings are done as shown above for DisplayName in SureFire/ANT task)

Add case key to Display name

Output: On import of the above report in AIO Tests, SCRUM-TC-2309 is updated. If case doesn’t exist then an appropriate error would be thrown

Mapping one automated case with multiple manual cases with JUnit5 @DisplayName

Add all case keys as part of the Display name

Output: On import of the above report in AIO Tests, 3 cases are updated with the result of the above execution → SCRUM-TC-2307, SCRUM-TC-2308 and SCRUM-TC-2350. If the case doesn’t exist then an appropriate error would be thrown

Mapping JUnit 4 case to existing manual case(s)

Method name can contain case key with underscores:

plantTree_should_returnNumberOfTrees_AT_TC_222.OR

The Automation Owner would have to mark the automated key value in an existing case to com.aio.tests.junit4.ForestCreatorTests.plantTree_should_returnNumberOfTrees

If a single automated JUnit case needs to update multiple cases, then the automated key needs to be updated in multiple AIO Tests automation key values.

Output: On import of the above report in AIO Tests, the automation key with com.aio.tests.junit4.ForestCreatorTests.plantTree_should_returnNumberOfTrees would be searched for and the cases marked with this key would be added to the cycle and updated.

Parametrized Named Cases Convention

With the release of the data-driven feature, it is now possible to create cases with datasets from JUnit cases.

The convention for creating data parameters with names is (params: parameterName1={0}, parameterName2={1}, parameterName1={2})

eg.

(params: userType={0},profile={1},income={2}). This example would create a case with 3 data parameters: userType, profile, and income and create datasets based on the Junit source values

The below example would create a test case with 5 datasets and a data parameter named month.

A new flag has been introduced updateDatassets. This flag controls, whether the change in datasets should be used to update the Junit datasets case or not. The value can be set to false, when a subset of datasets are being run to avoid removal of datasets. If set to true, this flag affects only existing datasets cases.

Parametrized Non-Named Cases

If the name is not specified, and a default report is generated, AIO Tests generates new parameter names for JUnit 5 reports, which elaborate on user data. However, on JUnit 4, in the absence of data in the report, cases would continue to work as individual cases.

In the below case, AIO Tests would create a parameter named Parameter 1 and would create 5 datasets with the numbers [1, 3, 5, -3, and 15] as values.

Output: The case with automation key of com.aio.tests.ParametrizedRunNonNamedTest.plantTree would be updated - 3 runs would be created with 1st and 3rd run showing pass and 2nd run showing Failed status

Importing Results

Post execution of a JUnit suite, the TEST-<xxx>.xml file can be uploaded either via

AIO Tests REST API calls using multipart form data to upload file
Cycle options in the Cycles Tab inside AIO Tests

Please follow the above links to continue to import results using either of the options.

For further queries and suggestions, please feel free to reach out to our customer support service via help@aiotests.com.