28 lines
861 B
Text
28 lines
861 B
Text
import algebra.ring data.nat
|
|
open algebra
|
|
|
|
variables {A : Type}
|
|
|
|
section
|
|
variable [s : group A]
|
|
include s
|
|
|
|
set_option blast.strategy "ematch"
|
|
|
|
attribute mul_one [forward]
|
|
attribute mul.assoc [forward]
|
|
attribute mul.left_inv [forward]
|
|
attribute one_mul [forward]
|
|
|
|
theorem inv_eq_of_mul_eq_one₁ {a b : A} (H : a * b = 1) : a⁻¹ = b :=
|
|
-- This is the kind of theorem that can be easily proved using superposition,
|
|
-- but cannot to be proved using E-matching.
|
|
-- To prove it using E-matching, we must provide the following auxiliary assertion.
|
|
-- E-matching can prove it automatically, and then it is trivial to prove the conclusion
|
|
-- using it.
|
|
-- Remark: this theorem can also be proved using model-based quantifier instantiation (MBQI) available in Z3.
|
|
-- So, we may be able to prove it using qcf.
|
|
assert a⁻¹ * 1 = b, by blast,
|
|
by blast
|
|
|
|
end
|