Since values change constantly in reports, your first
question may be "why write tests against values". This article
presents the idea that in cases where you are calculating fields, or doing data
dictionary lookups for fields, testing the values can validate the reports
internal "code".
The downside of testing for values is that this has to be
done against a data source. When we test the format, as we did in the first
article, there is not a lot performance issues. But when we need to go against
a data source, then our automated tests don't perform quickly. For this
reason, this article doesn't recommend that the test referenced in Part 2 be
included with your continuous integration builds. But any builds that go to
your testing environments are good candidates to put this in.
Also keep in mind that we need to make sure the data that
the report uses is consistent. We'll talk about techniques to do that.