Add hasPrecisionOf(int) assertion for BigDecimal

[armandino]

Check List:

• Fixes N/A
• Unit tests : YES
• Javadoc with a code example (on API only) : YES
• PR meets the contributing guidelines

AssertJ provides a hasScaleOf() assertion for the BigDecimal type. This PR adds a companion assertion hasPrecisionOf for verifying the BigDecimal.precision() method. I followed the hasScaleOf() implementation and tests in this PR.

[scordio]

Thanks @armandino! My comments are mostly due to the hasScale() / scale() patterns you probably followed. We would prefer to follow a different style for new assertions.

We might also improve hasScale() / scale() but this can be taken up in a separate issue or PR.

[scordio]

T seems to be unused, isn't it?

[armandino]

I tried removing it, but looks like it is required here:

  public AbstractBigDecimalPrecisionAssert<SELF> precision() {
    isNotNull();
    return new BigDecimalPrecisionAssert(myself);
  }

[scordio]

We briefly discussed this approach with the team.

We should add a bit of context to the error messages by showing the root object we navigated from.

For example, assuming:

assertThat(myBigDecimal).scale().isGreaterThan(5L)

If the assertion fails, it will show something like

expecting 2 to be greater than 5 but was not

where the fact that a BigDecimal precision was checked is forgotten.

Instead, the error message should have a better description and be more like:

[checking 2.123 precision] expecting 2 to be greater than 5 but was not

Unfortunately, we don't do this consistently in AssertJ so I didn't manage to find a good example quickly. I'll try to have another look in the next few days.

[armandino]

@scordio thanks for the feedback! It was very helpful. I addressed most of the comments, but there are still a few outstanding items.

[armandino]

Hi @scordio, I updated the PR. Please let me know if I missed anything.