Differential coverge numbers don't add up

[Vladyyy]

See the examples provided below:

I would expect that UBC + CBC = Total but that is not the case. In the example 243716 != 71615 + 171445 - > 243716 != 243060

Is my UBC + CBC = Total assumption wrong ?

[henry2cox]

Total should be equal to the sum of all the (non-deleted) coverpoints: UBC + CBC + LBC + + GBC + UIC + UNC + GNC +GIC That is: the sum of all the baseline coverpoints which still exist, plus all the new coverpoints created by newly added code as well as those created by changed compiler directives (say).

[Vladyyy]

Thanks, that makes sense.

A follow-up question: Does CBC / (CBC + UBC) * 100 represent the coverage percentage for the baseline

[henry2cox]

The short answer is "not quite":

• Covered baseline code includes CBC (covered in baseline and current), LBC (covered in baseline but not covered now), ECB (excluded code that was covered before) and DCB (deleted code which had been covered)
• Uncovered baseline code includes UBC, EUB, and DUB code - as well as GBC.
• as a result: the baseline covered ration is the sum of the latter divided by the sum of the latter plus former.

Of course, some of those categories may be empty - depending on what changed in your code and configuration.

This is discussed (to some degree) in the IEEE paper referred in the README - primarily in the section which talks about coverage criteria. The context of that discussion if how some more sophisticated module-, directory- or file-specifc criteria can be verified and enforced.

[Vladyyy]

Thanks, this was helpful.

his is discussed (to some degree) in the IEEE paper referred in the README - primarily in the section which talks about coverage criteria.

I did read the paper and watched you presentation