the after() function should execute after a failed test as well as after a passing test

[dianelaw]

If failfast is true, it probably doesn't matter if the after function executes, but when failfast is false, it's often more important to execute cleanup steps in the after function when a test fails than when it passes.

[dianelaw]

I did it by passing the after function to the function that checks failfast, prints the stack for failing tests, etc. Added the call just after the stack print

From run () (in runner .fs):
with
| ex when failureMessage <> null && failureMessage = ex.Message -> pass()
| ex -> fail ex test.Id suite.After
from fail (also in runner.fs):
let fail (ex : Exception) id after =
try
if failFast = ref true
then failed <- true
failedCount <- failedCount + 1
contextFailed <- true
let f = DateTime.Now.ToString("MMM-d_HH-mm-ss-fff")
let ss = screenshot configuration.failScreenshotPath f
reporter.fail ex id ss
after()

This would call it regardless of the value of failfast, You could call it conditionally as well.

[lefthandedgoat]

thanks I will work on getting something like this in tonight.

[dianelaw]

Jeremy just added his version as a pull request too. :-) Thanks, Chris!

[lefthandedgoat]

https://www.nuget.org/packages/canopy/0.9.25

[lefthandedgoat]

This has the fix in, 9.25 did not because I forgot to merge the change before building the package.

https://www.nuget.org/packages/canopy/0.9.26