Private val in value classes (Resolve TODO since 2.10 is dropped)

[barambani]

In relation to #2673 , this adds few more privates to constructor's parameters when extending AnyVal.
In the case of Either's CatchOnlyPartiallyApplied I added an exception to MiMa as there are good chances no one is really using dummy.

[barambani]

As a side note, I tried also working around the BC issues when doing the same in the Boilerplate generated syntaxes' Ops (that actually don't even extend AnyVal) but it's big work there. I will open a different PR if I find a reasonable way.

[kailuowang]

@barambani thanks very much. No need to work around BC issue for this one. We decided in #2617 that we should just add mima exceptions. If you want to be really sure, you can add a special mima test like this one https://github.com/typelevel/cats/pull/2614/files#diff-978c9b9bcdbb84476d4105d6abb2c151L21

[barambani]

I will try. The issue there is that, for the generated arity syntax functions, the MiMa exceptions needed are a lot, especially in 2.11.12. I will see if that's doable in a decent way. 👍

[codecov-io]

Codecov Report

Merging #2692 into master will decrease coverage by 0.27%.
The diff coverage is 0%.

@@            Coverage Diff             @@
##           master    #2692      +/-   ##
==========================================
- Coverage   95.13%   94.85%   -0.28%     
==========================================
  Files         365      365              
  Lines        6758     6760       +2     
  Branches      289      288       -1     
==========================================
- Hits         6429     6412      -17     
- Misses        329      348      +19

Impacted Files	Coverage Δ
...re/src/main/scala/cats/syntax/traverseFilter.scala	100% <ø> (ø)	⬆️
...laws/src/main/scala/cats/kernel/laws/package.scala	100% <ø> (ø)	⬆️
core/src/main/scala/cats/syntax/foldable.scala	100% <ø> (ø)	⬆️
core/src/main/scala/cats/syntax/reducible.scala	100% <ø> (ø)	⬆️
laws/src/main/scala/cats/laws/package.scala	100% <ø> (ø)	⬆️
core/src/main/scala/cats/syntax/either.scala	86.76% <ø> (-12.5%)	⬇️
...patTest/src/main/scala/catsBC/MimaExceptions.scala	0% <0%> (ø)	⬆️
testkit/src/main/scala/cats/tests/CatsSuite.scala	33.33% <0%> (-36.67%)	⬇️
.../src/main/scala/cats/data/RepresentableStore.scala	83.33% <0%> (-4.17%)	⬇️
... and 17 more

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 17293ed...46c0834. Read the comment docs.

[barambani]

I think this ready to be reviewed. Unfortunately I cannot change the generated TupleXSemigroupalOps to value classes or the MimaExceptions test

(1.asRight[String], 2.asRight[String], 3.asRight[String]) mapN (_ + _ + _)

wouldn't pass.

UPDATE: changed the test to

(1.validNel[String], 2.validNel[String], 3.validNel[String]) mapN (_ + _ + _)

[barambani]

Could the JVM 2.12's job be restarted ? Thanks

[kailuowang]

Thanks so much!

[kailuowang]

Quick question though, how come the changed test passed the mima test?

[barambani]

Sorry, lack of details from me. I changed the test just to make it more representative of the real life use cases. The change is only from Either to Validated as it makes more sense to use mapN on the latter. The situation didn't change. The boilerplate generated syntax for Semigroupal cannot be changed or it will fail.

[LukaJCB]

Thanks for working on this!