572751c56e
Before this commit, we were inferring whether an inductive/structure/class were meta or not. This was bad since the user had no clue whether the type was trusted (non meta) or not. Moreover, users could get confused by this behavior and assume the kernel was allowing trusted code to rely on untrusted one.
23 lines
310 B
Text
23 lines
310 B
Text
open tactic
|
|
|
|
attribute [instance]
|
|
meta def expr_to_app : has_coe_to_fun expr :=
|
|
{ F := λ e, expr → expr,
|
|
coe := expr.app }
|
|
|
|
meta constants f a b : expr
|
|
|
|
check f a
|
|
|
|
check f a b
|
|
|
|
check f a b a
|
|
|
|
set_option pp.coercions false
|
|
|
|
check f a b a
|
|
|
|
set_option pp.all true
|
|
set_option pp.coercions true
|
|
|
|
check f a b
|