Java assertj double containsSequence

Question

如何在比较双序列时指定对 assertj 的 double/float 容差？这里，items 是一个 List<ScoredItem> 和 ScoredItem 有 2 个属性 itemId 和 score.

assertThat(items).extracting("score").containsSequence(3.0, 2.0, 1.0, 0.35, 0.2);

单元测试期间抛出异常。

java.lang.AssertionError: 
Expecting:
 <[3.0, 2.0000000298023224, 1.0, 0.3500000052154064, 0.20000000298023224]>
to contain sequence:
 <[3.0, 2.0, 1.0, 0.35, 0.2]>

我也试过：

assertThat(items).extracting("score").containsSequence(new double[]{3.0, 2.0, 1.0, 0.35, 0.2}, Offset.offset(0.01));

...但运气不好:(

Answer 1

由于这是从列表中提取的，因此您必须在调用 containsSequence 之前使用 usingComparatorForType 调用通用方法（感谢纠正了我的颠倒顺序）。然后你可以传入 DoubleComparator.

assertThat(items)
   .extracting("score")
   .usingComparatorForType(new DoubleComparator(0.2), Double.class)
   .containsSequence(3.0, 2.0, 1.0, 0.35, 0.2);

如果items只是一个double []，那么你只需要调用方法usingComparatorWithPrecision

assertThat(items)
    .usingComparatorWithPrecision(0.5)
    .containsSequence(3.0, 2.0, 1.0, 0.35, 0.2);

或将第二行的 offset 替换为 withPrecision

assertThat(items).extracting("score")
    .containsSequence(new double[]{3.0, 2.0, 1.0, 0.35, 0.2}, withPrecision(0.5));

我已经使用 AssertJ 3.9.1 使用以下代码对此进行了测试：

public class ScoredItem {
    public int itemId;
    public double score;

    public ScoredItem(int i, double s) {
        itemId = i;
        score = s;
    }

    public static void main(String[] args) {
        List<ScoredItem> items = new ArrayList<ScoredItem>();
        items.add(new ScoredItem(1,3.0));
        items.add(new ScoredItem(1,2.0));
        items.add(new ScoredItem(1,1.1));
        items.add(new ScoredItem(1,0.33));
        items.add(new ScoredItem(1,0.22));

        assertThat(items)
            .extracting("score")
            .usingComparatorForType(new DoubleComparator(0.2), Double.class)
            .containsSequence(3.0, 2.0, 1.0, 0.35, 0.2);

    }
}

Answer 2

@Magdop 是正确的（断言后调用usingComparatorWithPrecision除外）。

使用 3.10.0，此测试通过：

@Test
public void containsSequence_with_precision() {
  // GIVEN
  double[] scores = { 3.0, 2.0000000298023224, 1.0, 0.3500000052154064, 0.20000000298023224 };
  // THEN
  assertThat(scores).containsSequence(new double[] { 3.0, 2.0, 1.0, 0.35, 0.2 }, withPrecision(0.01));
  assertThat(scores).usingComparatorWithPrecision(0.01).containsSequence(new double[] { 3.0, 2.0, 1.0, 0.35, 0.2 });
  assertThat(scores).usingComparatorWithPrecision(0.01).containsSequence(3.0, 2.0, 1.0, 0.35, 0.2);
  assertThat(scores).usingComparatorWithPrecision(0.0001).containsSequence(3.0, 2.0, 1.0, 0.35, 0.2);
  // fails as expected when uncommented
  // assertThat(scores).usingComparatorWithPrecision(0.0000000001).containsSequence(3.0, 2.0, 1.0, 0.35, 0.2);
}

usingComparatorWithPrecision 需要在断言之前调用，否则 AssertJ 不会使用它。

withPrecision 是 offset 的别名（恕我直言，更具可读性）。

Java assertj double containsSequence

Java assertj double containsSequence

java

unit-testing

floating-point

precision

assertj