to(Seq) doesn't evaluate collection when called on a Seq

[martijnhoekstra]

Fixes #236 Not so much about elegance as about getting the job done.

[NthPortal]

LGTM

Should LinearSeq be handled as well?

[martijnhoekstra]

@NthPortal yes, I guess

[martijnhoekstra]

I duplicated the logic for LinearSeq. The lack of elegance I was referring to earlier is starting to stick out even further.

stale

[martijnhoekstra]

I can't reconcile the documentation with the observation that the extractor object works, but the plain type match doesn't. Maybe there is a bug in the implementation.

…

On Sat, Jul 27, 2019, 19:47 Nth ***@***.***> wrote: ***@***.**** commented on this pull request. ------------------------------ In compat/src/main/scala-2.11_2.12/scala/collection/compat/CompatImpl.scala <#238 (comment)> : > import scala.collection.generic.CanBuildFrom import scala.collection.mutable.Builder import scala.collection.{immutable => i, mutable => m} +/* builder optimized for a single ++= call, which returns identity on result if possible + * and defers to the underlying builder if not. + */ +private final class IdentityPreservingBuilder[A, CC[X] <: TraversableOnce[X]](that: Builder[A, CC[A]])(implicit ct: ClassTag[CC[A]]) + extends Builder[A, CC[A]] { + var collection: CC[A] = null.asInstanceOf[CC[A]] + var ruined = false + + final override def ++=(elems: TraversableOnce[A]): this.type = + elems match { + case ct(ca) if (collection == null && !ruined) => { I guess it just doesn't work with HKT. It works fine with normal kinded types: Welcome to Scala 2.12.8 (OpenJDK 64-Bit Server VM, Java 1.8.0_212).Type in expressions for evaluation. Or try :help. scala> def test[A: reflect.ClassTag](elem: Any): Boolean = elem match { | case a: A => true | case _ => false | } test: [A](elem: Any)(implicit evidence$1: scala.reflect.ClassTag[A])Boolean scala> test[Int]("foo") res0: Boolean = false scala> test[Int](1) res1: Boolean = true scala> test[String](1) res2: Boolean = false scala> test[String]("foo") res3: Boolean = true — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#238?email_source=notifications&email_token=AAGXOEKA4Z74KEI3RQGFMMLQBSC2LA5CNFSM4IG2NCZ2YY3PNVWWK3TUL52HS4DFWFIHK3DMKJSXC5LFON2FEZLWNFSXPKTDN5WW2ZLOORPWSZGOB7YV5CY#discussion_r307972509>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AAGXOELU5RSW4W4NXVVYBXLQBSC2LANCNFSM4IG2NCZQ> .