Rescript formatting unstable and producing semantic deviations for inline functions in pipe chains

[mrmurphy]

Hi there! I'm upgrading my codebase to Rescript 11 and ran into an old file that has its semantic meaning changed when the formatter is run on it multiple times.

Original snippet:

let setProcessedRevocationForExParticipants = (
  ~tx=?,
  ~journalGID: GID.t,
  ~userIds: array<string>,
  ~timestamp: Js.Date.t,
  (),
) => {
  open DB2.SharedJournals
  update({"processed_revocation_at": timestamp})
  ->whereEq("journal_gid", journalGID->GID.toString)
  ->where(isNotNull("revoked_at"))
  ->where(isNull("processed_revocation_at"))
  ->(a => userIds->Js.Array2.length > 0 ? a->where(not_(in_(~col="with_user", userIds))) : a)
  ->DB.Main.Run.ignore(~tx?)
}

After saving the file once:

let setProcessedRevocationForExParticipants = (
  ~tx=?,
  ~journalGID: GID.t,
  ~userIds: array<string>,
  ~timestamp: Js.Date.t,
  (),
) => {
  open DB2.SharedJournals
  update({"processed_revocation_at": timestamp})
  ->whereEq("journal_gid", journalGID->GID.toString)
  ->where(isNotNull("revoked_at"))
  ->where(isNull("processed_revocation_at"))
  ->a => userIds->Js.Array2.length > 0 ? a->where(not_(in_(~col="with_user", userIds))) : a
  ->DB.Main.Run.ignore(~tx?)
}

Notice that the parens around (a => ...) have been dropped

And after another save:

let setProcessedRevocationForExParticipants = (
  ~tx=?,
  ~journalGID: GID.t,
  ~userIds: array<string>,
  ~timestamp: Js.Date.t,
  (),
) => {
  open DB2.SharedJournals
  update({"processed_revocation_at": timestamp})
  ->whereEq("journal_gid", journalGID->GID.toString)
  ->where(isNotNull("revoked_at"))
  ->where(isNull("processed_revocation_at"))
  ->(
    a =>
      userIds->Js.Array2.length > 0
        ? a->where(not_(in_(~col="with_user", userIds)))
        : a->DB.Main.Run.ignore(~tx?)
  )
}

Notice that the compiler now thinks that DB.Main.Run should be called on a directly, instead of on the result of the function that a belongs to.

Thank you for filing! Check list:

• Is it a bug? Usage questions should often be asked in the forum instead.
• Concise, focused, friendly issue title & description.
• A minimal, reproducible example.
• OS and browser versions, if relevant.
• Is it already fixed in master? Instructions

[mrmurphy]

Workaround for now is to extract that inline function into its own variable