If you interrupt AVA (Ctrl+C/SIGTERM) while tests are still pending, AVA should print those tests before exiting. Similarly it should print pending tests when --timeout is reached.
This should help debug which tests are taking too long.
Related to #171.
dflupu earned $80.00 by resolving this issue!
- Checkout the Issuehunt explorer to discover more funded issues.
- Need some help from other developers? Add your repositories on Issuehunt to raise funds.
Great idea!
:+1:
This would be awesome. 馃憤
Would be really useful to use why-is-node-running to also show why tests are still pending.
Just curious, has there been any movement on this issue?
cc: @novemberborn
@QuotableWater7 nope. Are you interested in taking this on?
This would be tremendously useful. Is this still something that is planned @novemberborn?
@austinkelleher most definitely. Would you want to give it a go?
Ran into this problem myself - gonna try hacking something together if y'all don't mind. 馃檪
This will be easier to implement now that #1722 has landed, though perhaps we should implement it in the reporters, in which case it'd be better to wait for https://github.com/avajs/ava/pull/1776.
any update on this now https://github.com/avajs/ava/pull/1722 is merged ? Would be very useful 馃槃
@vancouverwill I haven't had much time to work on #1776. I do think it should be implemented in the reporters, but I'm happy with some duplication if that unblocks.
Would you be interested in working on this?
@novemberborn hey yeah I would be interested, could you possibly give me some guidelines in terms of code style you would Iike or any goals with it? Thanks
@vancouverwill we use XO for code style.
The reporters should receive events when tests are declared and when they finish, so it'll be possible to keep a list of pending tests (and test files, too). Then in lib/cli there should be a hook for SIGINT & SIGTERM to make the reporters stop listening to incoming events and shut down the workers, and then print the pending tests & test files. The same for --timeout handling.
At least that's the theory, it may be a bit different in practice 馃槃
great, i've started to take a look at it, busy with the day job but hopefully I can get more time on the wk 馃槃
can you assign me so no one else ends up having to create duplicate/wasted work?
--verbose logger.We still want to implement this for when the process is interrupted ("killed"), and in the default mini reporter too.
Thank you @vancouverwill for getting us this far!
@issuehuntfest has funded $80.00 to this issue. See it on IssueHunt
@dflupu has submitted a pull request. See it on IssueHunt
@sindresorhus has rewarded $72.00 to @dflupu. See it on IssueHunt
Most helpful comment
Would be really useful to use
why-is-node-runningto also show why tests are still pending.