bound + recursion schemes

[csabahruska]

Is it possible to use bound with recursion-schemes library? i.e. to define the names in an AST using bound and doing AST transformations (optimisations, inlining, etc) with recursion-schemes

[Lysxia]

bound tracks the context of a term in the types, which changes when going through binders, this seems quite incompatible with the view in recursion-schemes of recursive types as fixed-points of a simple * -> * functor.

This QA also seems relevant https://dirtcheaphaskell.io/#2018-01-25-recursion-schemes-bound

[robrix]

I’m not using bound (yet), but I’ve had some recent success adapting Hinze’s approach from Efficient Generalized Folds:

https://github.com/robrix/path/blob/2da82c52cd17088fd6918ef3f37d3c13081dfd99/src/Path/Expr.hs#L32-L44

[freckletonj]

I tried a simple implementation of bound + recursion-schemes, ~and it works so far~

https://gist.github.com/freckletonj/c54a60085e225e5bc67af27dc4c31f37

~I'm not sure if it'll break in subtle ways, perhaps when changing bound variables during a fold? Or using a bound var type more involved than ()?~

EDIT: It cannot traverse into Scopes, oops, the pattern constructor for LamF wants a Scope ... not an r:

> :t PlusF
PlusF :: r -> r -> ExpF a r
> :t LamF
LamF :: Scope () Exp a -> ExpF a r

[phadej]

It cannot traverse into Scopes, oops, the pattern constructor for LamF wants a Scope ... not an r:

Exactly. It doesn't work correctly. With -ddump-splices

Test.hs:30:1-21: Splicing declarations
    makeBaseFunctor ''Exp
  ======>
    data ExpF (a_ablC :: ghc-prim-0.5.3:GHC.Types.Type) r_adrG
      = VF a_ablC |
        LamF (Scope () Exp a_ablC) |
        r_adrG :@$ r_adrG |
        IntF Int |
        PlusF r_adrG r_adrG
      deriving (Functor, Foldable, Traversable)

Note that in LamF we still have Exp a not r.