How to interpret "Stat" and "Off" values in C++Test coverage report?

spz1kor

I have the following code coverage report snippet and want to understand what do "Stat" and "Off" values represent and how to interpret them?

<CvgData cvgType="SC"> <Stat val="0;3;0;3;0;3;0"/> <Off val="0;76;115;120;146;153;198"/> </CvgData>undefined</File>undefined<File loc="/parasoft_example/parasoft_example/sub/sub.cpp"> <Content>int get_fifty() { return 50; } </Content>

Bogdan Czwartkowski

@spz1kor ,
I think that you are looking at incorrect piece of your XML report, or the code is not that simple. The CovData record that you shown must refer to another file.

A file as simple as this one
int get_fifty() \n {\n return 50;\n }
will only have three offsets - start of the file, start of the return statement and end of the return statement.
If I have a file with only these 4 lines:

int get_fifty()
{
    return 50;
}

then its statement coverage detail data will look like this:

            <CvgData cvgType="SC">
               <Stat val="0;3;0"/>
               <Off val="0;21;31"/>
            </CvgData>

Obviously, if there are comments before the function definition, the offsets will change naturally. if there is another coverable code before or after the function, there will be more offset values.

Bogdan Czwartkowski

@spz1kor ,

The <CovData> tags are added to XML reports when detailed coverage reports are enabled. They tell our HTML report generator which statements to highlight for coverage in the detailed HTML reports. The values in <Off> tag are byte offsets from the beginning of a source file, and <Stat> values represent coverage highlight (covered, non-covered, partial).

These tags are for our internal reporting and rendering purposes, I do not think they will be of any use for end users.

spz1kor

It would be helpful if you could elaborate the values in the example mentioned above. We need some custom reports and that info would be helpful.

Bogdan Czwartkowski

@spz1kor ,

I would like to stress again that these values are for coverage rendering in C/C++test's detailed coverage reports, hence it is all for internal use and it is subject to change. These values do not contain actual values of coverage achieved during tests. Still, here is an explanation given your example:

First of all, the starting XML tag is <File> and it has <Content> and <CovData> child tags which represent source file content and coverage data to render on the content, respectively.

Since <Content> is obvious, let us focus on <CovData>.
Given our example:
<CvgData cvgType="SC"> <Stat val="0;3;0;3;0;3;0"/> <Off val="0;76;115;120;146;153;198"/></CvgData>
we have:

• <CvgData cvgType="SC"> - "SC" means that details represent statement coverage, which was enabled for the detailed coverage report,
• <Off val="0;76;115;120;146;153;198"/> - represents places in the code specified as offsets (number of characters) counted from the beginning of the source file,
• <Stat val="0;3;0;3;0;3;0"/> - represents coverage state to render starting at particular offsets in the code, with the values meaning:
  • 0 - no coverage rendered (no highlight),
  • 1 - rendered as not covered (red),
  • 2 - partially covered (yellow),
  • 3 - fully covered (green).

spz1kor

@Bogdan Czwartkowski , thanks for the response, but I highly doubt if the "Off" values are correct here.
If you see the code snippet is very small, and we do have character offset as high as 76,115..198:

int get_fifty() \n {\n return 50;\n } -> total number of character bytes here is 39.

So these values really don't make sense in this context.

spz1kor

Makes sense!

spz1kor

Is there a way to get combined offset values for SC, DC and MCDC together?