It's really hard to detect the difference between actual and expected output right from the bats message when the test fails. Because bats doesn't indicate what the difference between expectation and real result:

Such a message doesn't really show what should be changed to make the test pass.
This issue is transferred from this repository.