How to Perform a Subfunction Unit Test (Batch)¶

Subfunction unit tests for batch processing are tested using an automated test framework. Tests are performed in subfunction units by continuously executing the request unit tests.

The test class should be created in such a way that the following conditions are met.

The package of the test class is the package of the subfunction to be tested.
Create a test class with the class name <subfunction ID> Test.

For example, if the subfunction ID of the target subfunction is B21AC01, the test class will be as follows.

package nablarch.sample.ss21AC01

import nablarch.test.core.batch.BatchRequestTestSupport;

// Middle is omitted

public class B21AC01Test extends BatchRequestTestSupport {

Test case division policy¶

Basically, one test case per sheet. Exceptions are shown below.

For complex test cases¶

When there is a large amount of test data, or when there are many processes included in one sheet, including all the test data into one sheet may result in too much data in the sheet decreasing the readability. Dividing one case into multiple sheets is better.

For very simple test cases¶

If the test case is very simple and the amount of test data is small, all test cases may be included in one sheet.

Basic description method¶

Basically, one test case is written in a single sheet. When multiple batch executions are written in a single sheet, it is a subfunction unit test.

In the following example, a subfunction consisting of three batches (file input batch, user deletion batch and file output batch) is executed.

/** Case that ends normally */
@Test
public void testSuccess() {
    execute();
}

[TestSuccess sheet]

LIST_MAP=testShots

no	description	expectedStatusCode	setUpTable	setUpFile	expectedTable	expectedFile	requestPath
1	File input	100	default	default	default		fileInputBatch
2	Delete user	100	default		default		userDeleteBatch
3	File Output	100	default		fileInputBatch	default	fileOutputBatch

When dividing a single test case into multiple sheets¶

For example, the test case illustrated in the previous section (Basic description method) can be divided and described as follows.

package nablarch.sample.ss21AA01

import org.junit.Test;
import nablarch.test.core.messaging.BatchRequestTestSupport;

// Middle is omitted

public class B21AA01Test extends BatchRequestTestSupport {

    @Test
    public void testSuccess() {

        // Register the input file to the temporary table
        execute("testSuccess_fileInput");

        // Delete the user-related table for temporary table information
        execute("testSuccess_userDelete");

        // Output the results to a file
        execute("testSuccess_fileOutput");
    }

[TestSuccess_fileInput sheet]

LIST_MAP=testShots

no	case	expectedStatusCode	setUpTable	setUpFile	requestPath
1	File input	100	default	default	fileInputBatch

[testSuccess_userDelete sheet]

LIST_MAP=testShots

no	case	expectedStatusCode	setUpTable	expectedTable	requestPath
1	Delete user	100	default	default	userDeleteBatch

[TestSuccess_fileOutput sheet]

LIST_MAP=testShots

no	case	expectedStatusCode	setUpTable	outFile	requestPath
1	File Output	100	default	default	fileOutputBatch

When multiple cases are included in one sheet¶

In the case of very simple test cases, they may be grouped together.

In the following example, two test cases (a normal case and a case with 0 records as input data) are described in a single sheet

/** Case that ends normally */
@Test
public void testSuccess() {
    execute();
}

[TestSuccess sheet]

LIST_MAP=testShots

no	description	expectedStatusCode	setUpTable	setUpFile	expectedTable	requestPath
1-1	File input	100	shot1	shot1		fileInputBatch
1-2	Delete user	100			shot1	userDeleteBatch
2-1	File input（0 record）	100	shot2	shot2		fileInputBatch
2-2	Delete user（0 record）	100			shot2	userDeleteBatch

Tip

Group IDs can be used to describe multiple cases of test data on one sheet. For more information, see the “To describe the data of multiple test cases on a single sheet” section.