I want to implement an algorithm in C# but test it in Python using python.net & pytest which I'm familiar with (and I also have a reference implementation in Python with which I want to compare outputs), so the question is: is there a way to compile C# to a DLL, import it in Python with python.net, run tests in Python and collect coverage of C# code being invoked during this process?
For example let's consider I have myclass.cs file
namespace mynamespace
{
public class MyClass
{
public static int mymethod(int x, int y)
{
return x + y;
}
}
}
after that I compile it with mcs
> mcs -t:library myclass.cs
getting myclass.dll which I import using python.net library in a binding.py
from pathlib import Path
import pythonnet
pythonnet.load()
from System import Reflection
Reflection.Assembly.LoadFile(str(Path(__file__).parent / 'myclass.dll'))
import mynamespace
mymethod = mynamespace.MyClass.mymethod
after that in my test.py
from binding import mymethod
def test_mymethod():
assert mymethod(1, 2) == 3
After running
> pytest test.py
I'm getting expected
...
test.py . [100%]
======================================================================= 1 passed in 0.41s ========================================================================
so far so good, but the question is how to get the coverage of original myclass.cs file? Is it even possible?
After searching I finally found a great dotnet-coverage tool which can be installed with
> dotnet tool install --global dotnet-coverage
and after that we can run Python tests with pytest (which call compiled C# code using python.net) and collect resulting coverage of C# project as easy as
> dotnet coverage collect pytest --output-format cobertura --output-format coverage.xml
which I can then for example upload to codecov service.