consider rename StringOps#lines. conflict jdk11 java.lang.String#lines

[xuwei-k]

• https://download.java.net/java/early_access/jdk11/docs/api/java.base/java/lang/String.html#lines()
• https://github.com/scala/scala/blob/v2.13.0-M5/src/library/scala/collection/StringOps.scala#L631

Welcome to Scala 2.13.0-M5 (Java HotSpot(TM) 64-Bit Server VM, Java 1.8.0_181).
Type in expressions for evaluation. Or try :help.

scala> "a".lines
res0: Iterator[String] = <iterator>

Welcome to Scala 2.13.0-M5 (Java HotSpot(TM) 64-Bit Server VM, Java 11-ea).
Type in expressions for evaluation. Or try :help.

scala> "a".lines
res0: java.util.stream.Stream[String] = java.util.stream.ReferencePipeline$Head@4730e0f0

[NthPortal]

what should it be replaced with? linesIterator? getLines (not in favour of this one myself)? iterateLines?

[lyricallogical]

Even if we change lines' signature, the risk still exists that the same problem will happen.
For solving it, we should add a method like ops for converting explicitly to StringOps.

[dwijnand]

Everyone wants to define lines on String - we've been here before: #5994 😄

[retronym]

In scala/scala-dev#557, I suggested to restore linesIterator as an alias for lines. People using that can cross build between 2.12 and 2.13.

https://github.com/retronym/scala/tree/topic/linesIterator

WDYT?

[retronym]

Marking as a blocker as it can't be done in a minor release.

[xuwei-k]

https://github.com/scala/scala/pull/7218/files#diff-a7e7b847970727f538a3c4f0416d36b8

[SethTisue]

.linesIterator cross-compiles, but gives a deprecation warning. a workaround that cross-compiles but involves nothing deprecated is .linesWithSeparators.map(_.stripLineEnd). (I know: yuck.)

[retronym]

A workaround, for those who don't need to use new parts of the Java standard library, is to use scalac -release 8.

$ java -version; scalac-launch 2.12.6 sandbox/test.scala; scalac-launch 2.12.6 -release 8 -Xprint:typer sandbox/test.scala
java version "11-ea" 2018-09-25
Java(TM) SE Runtime Environment 18.9 (build 11-ea+15)
Java HotSpot(TM) 64-Bit Server VM 18.9 (build 11-ea+15, mixed mode)
sandbox/test.scala:2: error: value toList is not a member of java.util.stream.Stream[String]
  "".lines.toList
           ^
one error found
[[syntax trees at end of                     typer]] // test.scala
package <empty> {
  class C extends scala.AnyRef {
    def <init>(): C = {
      C.super.<init>();
      ()
    };
    scala.Predef.augmentString("").lines.toList
  }
}

I still 👍 for undeprecating linesIterator.

[SethTisue]

(in the workaround ugliness-level sweepstakes, adding Predef.augmentString(...) beats mine on fewer characters, but loses on not being insertable in the middle of a chain of method calls)

[adriaanm]

un-deprecating linesIterator in scala/scala#7240, which we'll include in 2.12.7

[SethTisue]

note that it isn't safe to rely on the compiler to catch bad uses of .lines, you want to search your source code and find all of them

yes, often the compiler will catch it, e.g. s.lines.toVector fails to compile because Java's Stream doesn't have toVector

but more devious is something like s.lines.filter(...) == .... Java's Stream has filter, and == isn't type-safe, so if you're lucky you'll get a "are unrelated: they will most likely never compare equal" warning ... if you're lucky... and if you're lucky enough to spot the warning in your code's big pile of warnings it already generates that you tune out now...

I guess a Scalafix rewrite could help here, and wouldn't have false positives if you are using .lines method on something other than a String

[SethTisue]

gah, we removed .linesIterator entirely from 2.13. so .linesIterator won't become a cross-buildable alternative until we bring it back in 2.13.0-RC1 😢

it's okay though, Predef.augmentString(...).lines is fine. I'm just annoyed that I forgot about the removal in 2.13.