Like suggested here, I switched to the new Java-based sdf2table. However, this changed my ASTs completely:

Before:

Query(
  []
, SelectClause(
    None()
  , ExpAsVars([ExpAsVar(COUNT(None(), Star()), None())])
  )
, None()
, GraphPattern(
    [SingleVertex(ElemContents(Some("n"), None(), None(), None()))]
  )
, Some(WhereClause(Eq(False(), Not(True()))))
, None()
, None()
, None()
, None()
)

After:

Query(
  []
, SelectClause(
    (83, 69, 76, 69, 67, 84)
  , None()
  , ExpAsVars(
      [ ExpAsVar(
          COUNT(
            (99, 111, 117, 110, 116)
          , 40
          , None()
          , Star()
          , 41
          )
        , None()
        )
      ]
    )
  )
, None()
, GraphPattern(
    (77, 65, 84, 67, 72)
  , [SingleVertex(ElemContents(Some("n"), None(), None(), None()))]
  )
, Some(
    WhereClause(
      (87, 72, 69, 82, 69)
    , Eq(
        False((102, 97, 108, 115, 101))
      , Not(
          (78, 79, 84)
        , True((116, 114, 117, 101))
        )
      )
    )
  )
, None()
, None()
, None()
, None()
)

Is this supposed to happen? What are the numbers for?

Submitted by Oskar van Rest on 13 February 2018 at 19:54

Looks like these encode the ASCII characters of keywords?

Nice catch :-)
I guess the new Java implementation doesn’t handle the case-insensitive annotation yet.

That’s correct; I opened this issue last year: https://yellowgrass.org/issue/Spoofax/227

In that case I’m closing this issue (duplicates https://yellowgrass.org/issue/Spoofax/227).