Alleged duplicate result column names in 2.2

[area]

I just upgraded from 2.1.7 to 2.2, and I've found an apparent regression in some cypher queries. I have only managed to test on 2.1.5, but I believe I saw this behaviour on 2.1.7 also. The following query works in 2.1.5, using the Movie stars sample data.

MATCH (p:Person {name:"Tom Hanks"}) with p, 10 as prop RETURN sum(p.born)*prop as a, sum(p.born)*prop as b

But on 2.2, I get the error

Multiple result columns with the same name are not supported (line 1, column 75 (offset: 74))
"MATCH (p:Person {name:"Tom Hanks"}) with p, 10 as prop RETURN sum(p.born)*prop as a, sum(p.name)*prop as b"

I note that this error remains even if the field being sum is different in these two cases. For example:

MATCH (p:Person {name:"Tom Hanks"}) with p, 10 as prop RETURN sum(p.born)*prop as a, sum(p.name)*prop as b

I know that such a query makes no sense with this dataset, but it is sufficient to show that on 2.2.0, you get the same 'multiple result columns' error, and on 2.1.5 you get

SUM(Property(p,name(3))) can only handle numerical values, or null.

i.e. indicating this query has still passed the check for duplicate result columns.

The query

MATCH (p:Person {name:"Tom Hanks"}) with p RETURN sum(p.born) as a, sum(p.born) as b

returns correctly.

[cleishm]

Thanks for the report! I've confirmed the issue and will pass it along to our team.

[spacecowboy]

Fixed in #4340