running format with RedundantBraces rule twice produces different results after upgrade to 3.8.5 (from 3.8.3)

[visma-alexander-maslov]

After upgrading to 3.8.5 with RedundandBraces rule enabled and scala3 dialect.

Following code

  "Service" must {
    "have at least one test" in {
      pending
      // Write here your test ....
      // don't forget to remove `pending`
    }
}

is formatted to

"Service" must {
  "have at least one test" in
    pending
    // Write here your test ....
    // don't forget to remove `pending`
}

Which is kind of nice and correct.
But running scalaftm again formats it once more:

"Service" must {
  "have at least one test" in
    pending
  // Write here your test ....
  // don't forget to remove `pending`
}

The indentation for comments is removed.

[kitbellew]

please provide the information requested in the issue template.

[visma-alexander-maslov]

Where can I find a template ?
Pressing New Issue provides no template - just empty text field.

[tgodzik]

Looks like the old drafter broke, and we need to switch to the same thing as in metals https://github.com/scalameta/metals/tree/main/.github/ISSUE_TEMPLATE

[visma-alexander-maslov]

At least some issue discovered 😆 .. and fixed 👍
I'll try to go through that.

[kitbellew]

Looks like the old drafter broke, and we need to switch to the same thing as in metals https://github.com/scalameta/metals/tree/main/.github/ISSUE_TEMPLATE

yep, this was unfortunate. now we have a useless drop-down with a single Bug report template.

[visma-alexander-maslov]

Requeired: Configuration

version = "3.8.5"
runner.dialect = scala3
rewrite.rules = [  "RedundantBraces" ]

Required: Command-line parameters

scalafmt ./MySpec.scala

Steps

Given code like this:

import org.scalatest.BeforeAndAfterAll
import org.scalatest.wordspec.AnyWordSpec

class MySpec extends AnyWordSpec with BeforeAndAfterAll {

  "Service" must {
    "have at least one test" in {
      pending
      // Write here your test ....
      // don't forget to remove `pending`
    }
  }

  override protected def afterAll(): Unit = {
    println("After all....")
    super.afterAll()
  }

}

run scalafmt ./MySpec.scala 2 times.

Problem

Scalafmt formats code like this:

import org.scalatest.BeforeAndAfterAll
import org.scalatest.wordspec.AnyWordSpec

class MySpec extends AnyWordSpec with BeforeAndAfterAll {

  "Service" must {
    "have at least one test" in
      pending
    // Write here your test ....
    // don't forget to remove `pending`
  }

  override protected def afterAll(): Unit = {
    println("After all....")
    super.afterAll()
  }

}

Expectation

I would like the formatted output to look like this:

import org.scalatest.BeforeAndAfterAll
import org.scalatest.wordspec.AnyWordSpec

class MySpec extends AnyWordSpec with BeforeAndAfterAll {

  "Service" must {
    "have at least one test" in
      pending
      // Write here your test ....
      // don't forget to remove `pending`
  }

  override protected def afterAll(): Unit = {
    println("After all....")
    super.afterAll()
  }

}

Notes

The expected output produces if you run scalafmt ./MySpec.scala once.
Second time it seems to treat comments not belonging the the same indentation block.