On 03/23/2018 03:13 PM, Jeremy Korwin-Zmijowski wrote:
> I am using the SRFI-64 to build that test harness for my software. I evaluate the (load "main-test.scm") expression in the REPL to run my tests. By doing so, I can see the number of "expected passes" incrementing over the last runs.
>
> xxxxxx@(guile-user)> (load "main-test.scm")
> %%%% Starting test main
> # of expected passes 2
> xxxxxx@(guile-user)> (load "main-test.scm")
> %%%% Starting test main
> # of expected passes 4
> xxxxxx@(guile-user)> (load "main-test.scm")
> %%%% Starting test main
> # of expected passes 6
>
> I would like to have this behavior :
>
> xxxxxx@(guile-user)> (load "main-test.scm")
> %%%% Starting test main
> # of expected passes 2
> xxxxxx@(guile-user)> (load "main-test.scm")
> %%%% Starting test main
> # of expected passes 2
> xxxxxx@(guile-user)> (load "main-test.scm")
> %%%% Starting test main
> # of expected passes 2

I agree that would be more natural.

The specification of test-end says:

     Additionally, if the matching test-begin installed a new test-runner, then
     the test-end will de-install it, after reporting the accumulated test
     results in an implementation-defined manner.

So this seems like a bug.

I'm evaluating the attached patch.
--
	--Per Bothner
xxxxxx@bothner.com   http://per.bothner.com/