lean4-htt/tests/elab/5634.lean
Garmelon 08eb78a5b2
chore: switch to new test/bench suite (#12590)
This PR sets up the new integrated test/bench suite. It then migrates
all benchmarks and some related tests to the new suite. There's also
some documentation and some linting.

For now, a lot of the old tests are left alone so this PR doesn't become
even larger than it already is. Eventually, all tests should be migrated
to the new suite though so there isn't a confusing mix of two systems.
2026-02-25 13:51:53 +00:00

114 lines
2.3 KiB
Text
Raw Blame History

This file contains ambiguous Unicode characters

This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.

/-!
# Make sure `simpa` checks for metavariables in `using` clause
https://github.com/leanprover/lean4/issues/5634
-/
set_option linter.unnecessarySimpa true
/-!
This used to have a "don't know how to synthesize placeholder" error on the `have` line too.
This is because `have` is `refine_lift have ...; ?_`, so it indeed had a placeholder.
-/
/--
error: don't know how to synthesize placeholder for argument `a`
context:
htrue : True
⊢ False
---
error: unsolved goals
htrue : True
⊢ False
-/
#guard_msgs in
example : False := by
have htrue : True := trivial
simpa using id _
/-!
Simplified version of the test.
-/
/--
error: don't know how to synthesize placeholder for argument `a`
context:
⊢ False
---
error: unsolved goals
⊢ False
-/
#guard_msgs in
example : False := by
refine ?_
simpa using id ?_
/-!
Verifying that unassigned metavariables are OK, so long as they come from before elaboring the `using` clause.
-/
example (p : Prop) (h : p) : p := by
have : ?a := ?b
simpa using ?b
exact h
/-!
Occurs check
-/
/--
error: Occurs check failed: Expression
?foo
contains the goal ?foo
-/
#guard_msgs in
example : False := by
refine ?foo
simpa using ?foo
/-!
Regression test: need to synthesize postponed metavariables before metavariable checks.
-/
example (α : Type) (x : α) : Nonempty α := by
simpa using ⟨x⟩
/-!
Regression test: elaborates implicit arguments in the `using` clause
-/
noncomputable example (α : Type) [Nonempty α] : α := by
simpa using fun {β : Type} [inst : Nonempty β] => Classical.choice inst
/-!
Regression test: elaborates using implicit lambda feature
-/
example (p : Prop) (h : p) : ∀ {_ : Nat}, p := by
simpa using h
/-!
Regression test: make sure `simpa?` reports lemmas for both the goal and the `using` clause
-/
/--
info: Try this:
[apply] simpa only [id] using h
-/
#guard_msgs in example (p : Prop) (h : p) : id p := by
simpa? only [id] using h
/--
info: Try this:
[apply] simpa only [id] using h
-/
#guard_msgs in example (p : Prop) (h : id p) : p := by
simpa? only [id] using h
/-!
Regression test: unnecessary `simpa`
-/
def foo (n : α) := [n]
/--
warning: Try `simp at h` instead of `simpa using h`
Note: This linter can be disabled with `set_option linter.unnecessarySimpa false`
-/
#guard_msgs in
example (h : foo n ≠ [n]) : False := by
simpa [foo] using h