Unused Lit.Byte and Lit.Short

[MasseGuillaume]

I suppose this is for Typelevel: https://github.com/typelevel/scala/pull/12/files#diff-0023b3bfa053fb16603156b785efa7adR1011

0z // Lit.Byte
0s // Lit.Short

However:

import scala.meta._ 
import scala.meta.dialects.Typelevel212 

"object A { 0s }".parse[Source]

Error(
  Range(String("object A { 0s }"), 11, 11),
  "Invalid literal number",
  TokenizeException(
    Range(String("object A { 0s }"), 11, 11),
    "Invalid literal number"
  )
)

"object A { 0z }".parse[Source]

Parsed[Source] = Error(
  Range(String("object A { 0z }"), 11, 11),
  "Invalid literal number",
  TokenizeException(
    Range(String("object A { 0z }"), 11, 11),
    "Invalid literal number"
  )
)

[xeno-by]

These trees were introduced before Typelevel Scala compiler added support for byte and short literals (typelevel/scala#12). The main reason was consistency with other literals and needing to have data structures that represent publicly visible bits of internal compiler state (e.g. inferred types of final vals).

But yeah, this is inconsistent with the current reality in which TLS supports those literals. We should update the parser to accommodate that functionality.

[tgodzik]

Typelevel compiler is now archived: https://github.com/typelevel/scala