MuhammadTahaNaveed commented on issue #829:
URL: https://github.com/apache/age/issues/829#issuecomment-1519109605
> Ideally, the second MATCH should work as an intersection over the results
of the first MATCH. So the result should be only one vertex " {"id":
844424930131969, "label": "Person", "properties": {"name": "bobby"}}::vertex".
This is how it works in Neo4j as well, and the second MATCH does perform
intersection in case of properties with AGE as well. As demonstrated here
>
> ```
> test=# SELECT * from cypher('test', $$
> CREATE ({name:'bobby', age:25}),
> (:Person {name:'bobby'}),
> (:Bobby {name:'bobby', age:25, cigar:2})
> $$) as (a agtype);
> a
> ---
> (0 rows)
>
> test=# SELECT * from cypher('test', $$
> MATCH (u {name:'bobby'})
> MATCH (u {cigar:2})
> RETURN u
> $$) as (a agtype);
> a
>
------------------------------------------------------------------------------------------------------------
> {"id": 1407374883553282, "label": "Bobby", "properties": {"age": 25,
"name": "bobby", "cigar": 2}}::vertex
> (1 row)
> ```
>
> Another interesting result is this
>
> ```
> test=# SELECT * from cypher('test', $$
> MATCH (u:Bobby {name:'bobby'})
> MATCH (u:Person)
> RETURN u
> $$) as (a agtype);
> a
>
------------------------------------------------------------------------------------------------------------
> {"id": 1407374883553281, "label": "Bobby", "properties": {"name":
"bobby"}}::vertex
> {"id": 1407374883553282, "label": "Bobby", "properties": {"age": 25,
"name": "bobby", "cigar": 2}}::vertex
> (2 rows)
> ```
>
> And this
>
> ```
> test=# SELECT * from cypher('test', $$
> MATCH (u:Bobby {name:'bobby'})
> MATCH (u:Person {age:25})
> RETURN u
> $$) as (a agtype);
> a
>
------------------------------------------------------------------------------------------------------------
> {"id": 1407374883553282, "label": "Bobby", "properties": {"age": 25,
"name": "bobby", "cigar": 2}}::vertex
> (1 row)
> ```
>
> It appears that AGE performs filtration based on the first MATCH clause,
ignores the label (if any) of the second clause, only performs filtration based
on the given properties and returns the result.
>
> It should instead give an error regarding support of multiple labels (as
of now).
>
> Hence, yes, I would call it a bug.
>
> Given that in case 2 labels are provided in 2 different MATCH clauses, it
does not perform an intersection of the 2 MATCH clauses, instead completely
ignores the label provided in the second MATCH clause.
Thanks for your valuable response.
In addition to your response, the below query also filters out results in
neo4j
`MATCH (u),(u:Person) RETURN u`
But this same query errors out in AGE saying `ERROR: multiple labels for
variable 'u' are not supported`
So the query
`MATCH (u) MATCH (u:Person) RETURN` should also error out the same until it
is supported.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: dev-unsubscr...@age.apache.org
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
