14 May
2009
14 May
'09
5:40 p.m.
Hi and thanks,
This is true- we can use warn to mean aborting test although it seems to go against what most "first readers" of the code would think it is.
warning by itself does not cause test case to abort.
That is clear- again my concern is with the workflow that is triggered after execution is complete and people are reading the test reports. A warn coupled with a logging message that says test was aborted should be clear enough. BOOST_REQUIRE won't work in the setup step because it does exactly what I want to avoid :) Thanks, I will throw a c++ exception. Greg