You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
(typing gamma v tau && stlc_abs_ty v ty && stlc_abs_body v body
&& stlc_tyctx_hd gamma1 ty && stlc_tyctx_tl gamma1 gamma)
#==> (typing gamma1 body body_ty)
Does not have any constraints on what body_ty can be and is unsound.
I would propose the following axiom instead, which I've checked against my coq definition.
let[@axiom] stlc_abd_typing_rev (gamma : stlc_tyctx) (v : stlc_term)
(tau : stlc_ty) (ty : stlc_ty) (body : stlc_term) (body_ty : stlc_ty)
(gamma1 : stlc_tyctx) =
(typing gamma v tau && stlc_abs_ty v ty && stlc_abs_body v body
&& stlc_ty_arr1 tau ty && stlc_ty_arr2 tau body_ty && stlc_tyctx_hd gamma1 ty
&& stlc_tyctx_tl gamma1 gamma)
#==> (typing gamma1 body body_ty)
The text was updated successfully, but these errors were encountered:
The following axiom
underapproximation_type/data/predefined/axioms.ml
Lines 497 to 502 in 94d0f05
Does not have any constraints on what
body_ty
can be and is unsound.I would propose the following axiom instead, which I've checked against my coq definition.
The text was updated successfully, but these errors were encountered: