Google Test ASSERT_ macros not modelled ideally

[Timothy Brackett]

assert is not a particularly appropriate model for Google Test's ASSERT_ macros. ASSERT_ macros are evaluated in Release builds, so the diagnostic assertWithSideEffect does not reasonably apply.

I don't know what the ideal model would be, but at least the model used for EXPECT_ macros would avoid the false-positive assertWithSideEffect findings.

[Daniel Marjamäki]

Thanks! Sounds reasonable. I guess something that uses throw/exit would be a better idea. If you have time to look into this feel free to provide a github pull request.