Asserting implication sets premise equalities

[RandomActsOfGrammar]

An assertion with an equality to the left of an implication results in the two sides being equal for the remainder of the proof.

Consider the following theorem:

Theorem error : forall (x y a b : o), a = b.
intros. assert (a = b -> x = y).
skip.
search.

The state after the assert is

Subgoal 1:

Variables: x y b
============================
 b = b -> x = y

Subgoal is:
 b = b

Note that the second subgoal has been changed from a = b to b = b.

We get the same behavior if we put the implication under a disjunction. In the following theorem

Theorem error2 : forall (x y a b : o), a = b.
intros. assert (x = y \/ (a = b -> x = y)).
skip.
search.

the state after the assert is

Subgoal 1:

Variables: x y b
============================
 x = y \/ (b = b -> x = y)

Subgoal is:
 b = b

However, if we instead put it under a conjunction, the two sides do not become equal. In proving

Theorem fine : forall (x y a b : o), a = b.
intros. assert (x = y /\ (a = b -> x = y)).
skip.
skip.

the proof state after the assert is

Subgoal 1:

Variables: x y a b
============================
 x = y /\ (a = b -> x = y)

Subgoal is:
 a = b

as we expect it to be.

I ran all these examples using Abella 2.0.6.

[JimmyZJX]

I believe this is what #113 describes, which is fixed in the master branch. Sadly the fix is not yet released.

[RandomActsOfGrammar]

I think you're right. I didn't find that one when I was looking for an existing issue.