* fix : make `mk_no_confusion_type` handle delta-reduction when checking the inductive type. * tests: extend `2500.lean`
15 lines
533 B
Text
15 lines
533 B
Text
/-!
|
|
# Verify that the `noConfusion` lemma succeeds at being generated, despite the inductive type not being a syntactical telescope
|
|
Fixes https://github.com/leanprover/lean4/issues/2500
|
|
Ensures the `to_telescope` call in `mk_no_confusion_type` (`src/library/constructions/no_confusion.cpp`) takes the global environment into account,
|
|
thus allowing for delta-reduction.
|
|
-/
|
|
|
|
def family := Type → Type
|
|
|
|
inductive bad : family
|
|
|
|
def Set (A : Type _) := A → Prop
|
|
|
|
inductive Thing (s : Set V) : Set V
|
|
| basic : ∀ x, s x → Thing s x
|