Closed bajtos closed 8 years ago
Looks good so far, I agree with the idea of testing this more thoroughly. And it does indeed appear 34b0665 broke some backcompat, but you could argue previous behavior was a bug in the first place. Still worthwhile for cleaning up for good in 3.0.
See also https://github.com/strongloop/strong-remoting/pull/265 and especially https://github.com/strongloop/strong-remoting/pull/265#issuecomment-164527886 as handling this code was a comedy of errors - somebody inside strongloop force-pushed it out without any notice, then only half the PR was re-merged and released without review, causing significant breakage as the code was incomplete (it was two commits and should not have been merged incomplete).
Done. @STRML @ritch PTAL
I have cleaned up the code a lot, restructured test cases and added comments to make it more clear what is being tested. The patch is ready for another round of code review.
@STRML @ritch PTAL.
The high-level overview of changes between 2.16.x versions can be found here: https://gist.github.com/bajtos/8afaa9d1ed7d7b410c76ec2095a8a992
LGTM
I don't see anything that should change. But there is quite a bit I may be overlooking still. I think the size of this test suite points to a separate module making sense for coercion.
The documentation in these tests would also be helpful to everyone. I think we should create a section about coercion in the readme that lists out the nuances of coercion. Which can be done outside of this PR.
@slnode test please
I did not see any problem either. The patch LGTM.
LGTM too should be really helpful for us to prevent regression on a much more detailed level :+1:
@ritch @Amir-61 @davidcheung thank you for the review!
The documentation in these tests would also be helpful to everyone. I think we should create a section about coercion in the readme that lists out the nuances of coercion. Which can be done outside of this PR.
I filled a new task for that - https://github.com/strongloop/strong-remoting/issues/313, I think it will be best to document the new behaviour after we make the intended changes/fixes.
An integration test suite describing and verifying argument coercion in REST.
(see https://github.com/strongloop-internal/scrum-loopback/issues/706#issuecomment-217371058 for more information)
In this pull request, I am describing the current behaviour in the
master
branch. My intention is to run the test suite against different strong-remoting versions to find differences, I'll post results later.Connect to strongloop-internal/scrum-loopback#884 Connect to strongloop-internal/scrum-loopback#913
@ritch @STRML please take a look