Test cleanup

[calebwin]

I want to be able to run some cleanup code after my tests run regardless of whether they succeed or fail. I though I could just do this by wrapping my call to runtests with a try... finally... but it seems like when an error occurs, the whole thing terminates and my cleanup code never gets run.

Any suggestions on how to intercept failures like this?

[rfourquet]

I don't know a way of doing that currently. Do you have a way with the regular Test library? It's a feature I've been planning and which shouldn't be difficult to implement, but the API was no totally setlled in my mind.