typelevel/discipline

Carlos Quiroz

@cquiroz

but on the latest release it requires me to include a checkAll method

Thus I can’t use it for cases besides law checking

Why was it changed?

Ross A. Baker

@rossabaker

What trait are you extending?

Are you using the scalatest integration?

Carlos Quiroz

@cquiroz

I’m extending CatsSuite with scalatest

Ross A. Baker

@rossabaker

checkAll became abstract so that more than AnyFunSuiteLike could be supported.

There are a few implementations in the same package as Discipline, if you've picked a scalatest style, that define checkAll. You might want one of those.

Carlos Quiroz

@cquiroz

it seems I may néd to update cats-testkit

CatsSuite extended Discipline but now it extends FunSuiteDiscipline

Ross A. Baker

@rossabaker

:+1:

Raúl Piaggio

@rpiaggio

In case anyone's interested, I made https://github.com/rpiaggio/discipline-munit. It's quite trivial given MUnit's integration with Scalacheck, but it bridges the gap. Current version is 0.1.0. Please let me know if you want me to add a PR mentioning it in discipline's README.

Ross A. Baker

@rossabaker

@rpiaggio :+1: That'd be great.

Raúl Piaggio

@rpiaggio

@rossabaker Thanks, I will do that. BTW, is it OK to use the org.typelevel.discipline.munit package name for this even though the lib is not under Typelevel? The reason for this is to be consistent with with the specs2 and scalatest integrations.

Ross A. Baker

@rossabaker

@rpiaggio Would you want to just move it to Typelevel?

Then it'd be consistent in package and location. Might help people find it, and of course we could leave you as an admin on it.

Raúl Piaggio

@rpiaggio

@rossabaker yes, I'd be happy to.

Ross A. Baker

@rossabaker

Great! I've only ever transferred repos to orgs I could already create repos in. I'm not sure if you can try to transfer, and then it triggers a notification.

When you transfer a repository that you own to another user account, the new owner will receive a confirmation email. The confirmation email includes instructions for accepting the transfer. If the new owner doesn't accept the transfer within one day, the invitation will expire.
The original owner of the repository is added as a collaborator on the transferred repository. Other collaborators to the transferred repository remain intact.

That looks like exactly what we want. Gets the project alongside its siblings, and keeps you in charge of it.

The maintenance on the others has mostly been merging scala-steward PRs. :laughing:

Raúl Piaggio

@rpiaggio

OK, I'll do the transfer, thank you! but before that, I have the dual question as before since I'll probably implement the suggestion from https://gitter.im/scalameta/munit?at=5eb3d132d6b5a360e43c339a. So, is it OK to move the project into Typelevel if the package name is not org.typelevel.discipline.munit anymore, but just munit ?

Ross A. Baker

@rossabaker

Oh, I think that's fine, too.

I am an old Java programmer, and I like the reverse domain name convention, and I like it when the organization ID matches the package name. But these opinions all make me old-fashioned.

There are other Typelevel projects that have a bare root name. jawn comes to mind.

Raúl Piaggio

@rpiaggio

:+1: thank you

Raúl Piaggio

@rpiaggio

I can't do the transfer:

To transfer a repository that you own to an organization, you must have permission to create a repository in the target organization.

You don’t have the permission to create public repositories on typelevel

Ross A. Baker

@rossabaker

Oh, hmm. The doc I found suggested that we'd get a notification on that to accept it.

Oh, I see the fine print now.

To transfer a repository that you own to an organization, you must have permission to create a repository in the target organization.

You could transfer it to another person, and they can accept. But you can't transfer it to an org.

Ross A. Baker

@rossabaker

The transferer has to be an owner, and I don't think a personal project can have more than one owner, so I don't think you can add me to your project.

Maybe you could transfer it to me, and then I could transfer it to Typelevel. Or maybe someone else has a better idea.

Sorry that this is a hassle. I thought it would be easy and make the project more visible.

I have a new project on discipline-specs2 that I'd like to move to munit as soon as the generator seed bug is resolved, so I'm excited to use this.

Raúl Piaggio

@rpiaggio

issue solved, @larsrh added me to typelevel. thank you! I just transferred the repo.

Raúl Piaggio

@rpiaggio

I guess I need some permission in sonatype to publish under org.typelevel, or someone else has to do the publishing

Ross A. Baker

@rossabaker

Oh, perfect!

Ross A. Baker

@rossabaker

To publish, you just need to open an issue on Sonatype's OSSRH JIRA.

Tag me, and I'll vouch for it, and they'll set you up.

And in the meantime, I can publish. But the Sonatype folks are pretty quick.

Raúl Piaggio

@rpiaggio

excellent, I'll open the issue

Raúl Piaggio

@rpiaggio

discipline-munit is published under org.typelevel now, starting with 0.2.0

Carlos Quiroz

@cquiroz

Thanks @rpiaggio

Ross A. Baker

@rossabaker

:tada: Looking forward to using it.

Amaal Ali

@amaalali

Is the plan for discipline-scalatest to support of other ScalaTest testing styles that aren't included at the moment (for example FreeSpec), or is it that users are being encouraged to implement a Discipline themselves for the style that they like?

Ross A. Baker

@rossabaker

@amaalali We added WordSpec a couple months ago. I don't see a particular reason we can't add others, as long as we're fairly confident scalatest will continue to support them. One thing we want from this library is binary stability.

Raúl Piaggio

@rpiaggio

Releasing discipline-munit 0.3.0. The minor version bump is due to the upgrade to Scalajs 1.2.0.

Philip Reiner Kensche

@vinjana

I have a question concerning one of the laws's tests in cats.kernel.laws. I already tried in typelevel/cats, without response, but it seems this channel here may be the better place. In EqLaws it says

def antiSymmetryEq(x: A, y: A, f: A => A): IsEq[Boolean] =
    (!E.eqv(x, y) || E.eqv(f(x), f(y))) <-> true

What is the purpose of the function f? Does it not just complicate things? Can I simply set this to identity? I searched for various definitions of antisymmetry, but could not find an additional function in any of them. If you have a hint where I can read about it, I'd welcome a tip.

Where communities thrive

People

Repo info

Activity