How to define display name using nested attributes of arguments in ParameterizedTest(name = #{index} multiply {0[0]} x {0[1]} = {0[2]})

Question

I cannot figure out if it is a missing feature, see the JUnit issue 1154 and my comment there, or just my inability to write properly the syntax for name of the @ParameterizedTest in JUnit5. Given the fact that issue has been open for last 3 years, I'm afraid I cannot wait till I get answer or even the implementation there, so I'm trying to ask here as well.

In my example ParameterizedTest, I'm using @MethodSource, which returns a Stream of ints arrays and I would like to use only those array's attributes instead of all {arguments} which is anyway same if I use {0}. Hard to explain verbally here, better to use code example bellow.

package net.delphym.unittests;

import org.junit.jupiter.api.DisplayName;
import org.junit.jupiter.params.ParameterizedTest;
import org.junit.jupiter.params.provider.MethodSource;

import java.util.stream.Stream;

import static org.junit.jupiter.api.Assertions.assertEquals;

class DynamicTestCreationTestParam {
    private static Stream<int[]> data() {
        return Stream.of(new int[][] { {1, 2, 2}, {5, 3, 15}, {121, 4, 484} });
    }

    @DisplayName("Multiplication test")
    @ParameterizedTest(name = "#{index} for {arguments}: {0} x {1} = {2}")
    @MethodSource("data")
    void testWithStringParameter(int[] data) {
        MyClass tester = new MyClass();
        int m1 = data[0];
        int m2 = data[1];
        int expected = data[2];
        assertEquals(expected, tester.multiply(m1, m2));
    }

    // class to be tested
    class MyClass {
        public int multiply(int i, int j) {
            return i *j;
        }
    }
}

The test results output looks something like this for the 1st run: #1 for [1, 2, 2]: [1, 2, 2] x {1} = {2}

Ideally, I would like to see this: #1 multiply: 1 x 2 = 2.

So question is what would be the @ParametrizedTest(name) definition to match my expectation?

I was trying something like this: name = "#{index} multiply : {[0][0]} x {[0][1]} = {[0][2]}" But it is syntacticly incorrect.

Answer 1

Alright, as suggested by @Murat Karagöz, I'm only reposting suggestions which I receive on Github to have all at one place.

In my opinion, without no doubt, those elegant solutions are only workarounds to the JUnit5 @ParameterizedTest API limitations, but they both satisfied my needs.

Proposed workaround #1 if I don't need to use int[]

Solution presented here is preserving @MethodSource("data") and only altering its return type. Now the data() method returns Stream of Arguments objects (3 in each), instead Stream of int arrays (of 3 in each).

Then the code reads:

private static Stream<Arguments> data() {
    return Stream.of(Arguments.of(1, 2, 2), Arguments.of(5, 3, 15), Arguments.of(121, 4, 484));
}

@DisplayName("Multiplication test")
@ParameterizedTest(name = "#{index}: {0} x {1} = {2}")
@MethodSource("data")
void testWithStringParameter(int m1, int m2, int expected) {
    MyClass tester = new MyClass();
    assertEquals(expected, tester.multiply(m1, m2));
}

Credits goes to Marc Philipp for his post @ Github.

Proposed workaround #2 to unroll test data to match display needs##

This solution provided by sormarus suggest to use @CsvSource instead of the @MethodSource and provides the set of the test data within the annotation.

I like this as simple example how to quickly prepare same basic variation without a need to have a specific data generating method.

The code looks like this:

    @DisplayName("Multiplication test")
    @ParameterizedTest(name = "#{index} multiply: {0} x {1} = {2}")
    @CsvSource({"1,2,2", "3,5,15", "121,4,484"})
    void testWithStringParameter(int m1, int m2, int expected) {
        MyClass tester = new MyClass();
        assertEquals(expected, tester.multiply(m1, m2));
    }