As of the latest release of the global tool (4.8.6), reports generated for generic classes are no longer accurate.
Console output of ReportGenerator:
2021-03-04T10:59:40: Arguments
2021-03-04T10:59:40: -reports:D:\GitSrc1\common\Common.Data\Coverage\Temp*\coverage.cobertura.xml
2021-03-04T10:59:40: -reporttypes:HtmlInline_AzurePipelines_Dark
2021-03-04T10:59:40: -targetdir:D:\GitSrc1\common\Common.Data\Coverage
2021-03-04T10:59:40: -assemblyfilters:-.Tests
2021-03-04T10:59:40: -filefilters:-*.g.cs
2021-03-04T10:59:42: Writing report file 'D:\GitSrc1\common\Common.Data\Coverage\index.html'
2021-03-04T10:59:43: Report generation took 2.9 seconds
It looks like _all_ classes with generic type parameters are affected by this. However, the below minimal class and test easily show the problem too:
public class GenericClass<T>
{
public string GetType()
{
return typeof(T).FullName;
}
}
using Xunit;
public class GenericClassTests
{
[Fact]
public void ItWorks()
{
Assert.Equal("System.Int32", new GenericClass<int>().GetType());
}
}
While the test does succeed and clearly covers the code, the generated report does report 0 uncovered lines, but also reports 0 covered lines which is clearly wrong.
Sorry for that.
I fixed it in release 4.8.7.
Most helpful comment
Sorry for that.
I fixed it in release 4.8.7.