Unexpected results when using GROUP BY GROUPING SETS and bind variables

Previous Topic Next Topic
 
classic Classic list List threaded Threaded
5 messages Options
Reply | Threaded
Open this post in threaded view
|

Unexpected results when using GROUP BY GROUPING SETS and bind variables

Aner Perez
I wrote a query for a report that needed to summarize table data for different subsets of an ARRAY column.  Once I had a working query, I recreated it in my code using an SQL query builder (the awesome jOOQ in this case).  Unfortunately the output from the generated SQL was different from the handwritten query.  Some of the grouping columns had null in them even though the original query showed the correct values.

After some help from the jOOQ users mailing list, I found that the generated query I wrote was inadvertently using bind variables for the array indexes used in the GROUPING SETS.  This apparently affects how the query is executed and leads to unexpected output.

I wrote an SQL script runnable in psql to demonstrate the issue.  The script creates a table with sample data and then executes 2 queries against it.  The first one has everything inlined and produces the expected output.  The second one uses a prepared statement and bind variables to reproduce the issue.

I'm not really sure if this is a bug in query validation or a bug in executing the query or if it's just unreasonable expectations on my part but the output of the 2 queries seems to violate the principle of least surprise.

- Aner

-- bind-group-by-test.sql --
CREATE TABLE bind_group_by (
    path text [] PRIMARY KEY,
    value int
);
INSERT INTO bind_group_by (path, value) VALUES
    (ARRAY ['A1', 'B1', 'C1', 'D1'], 0),
    (ARRAY ['A1', 'B1', 'C1', 'D2'], 1),
    (ARRAY ['A1', 'B1', 'C1', 'D3'], 2),
    (ARRAY ['A1', 'B1', 'C2', 'D1'], 3),
    (ARRAY ['A1', 'B1', 'C2', 'D2'], 4),
    (ARRAY ['A1', 'B1', 'C2', 'D3'], 5),
    (ARRAY ['A1', 'B1', 'C3', 'D1'], 6),
    (ARRAY ['A1', 'B1', 'C3', 'D2'], 7),
    (ARRAY ['A1', 'B1', 'C3', 'D3'], 8),
    (ARRAY ['A1', 'B2', 'C1', 'D1'], 9),
    (ARRAY ['A1', 'B2', 'C1', 'D2'], 10),
    (ARRAY ['A1', 'B2', 'C1', 'D3'], 11),
    (ARRAY ['A1', 'B2', 'C2', 'D1'], 12),
    (ARRAY ['A1', 'B2', 'C2', 'D2'], 13),
    (ARRAY ['A1', 'B2', 'C2', 'D3'], 14),
    (ARRAY ['A1', 'B2', 'C3', 'D1'], 15),
    (ARRAY ['A1', 'B2', 'C3', 'D2'], 16),
    (ARRAY ['A1', 'B2', 'C3', 'D3'], 17),
    (ARRAY ['A1', 'B3', 'C1', 'D1'], 18),
    (ARRAY ['A1', 'B3', 'C1', 'D2'], 19),
    (ARRAY ['A1', 'B3', 'C1', 'D3'], 20),
    (ARRAY ['A1', 'B3', 'C2', 'D1'], 21),
    (ARRAY ['A1', 'B3', 'C2', 'D2'], 22),
    (ARRAY ['A1', 'B3', 'C2', 'D3'], 23),
    (ARRAY ['A1', 'B3', 'C3', 'D1'], 24),
    (ARRAY ['A1', 'B3', 'C3', 'D2'], 25),
    (ARRAY ['A1', 'B3', 'C3', 'D3'], 26),
    (ARRAY ['A2', 'B1', 'C1', 'D1'], 27),
    (ARRAY ['A2', 'B1', 'C1', 'D2'], 28),
    (ARRAY ['A2', 'B1', 'C1', 'D3'], 29),
    (ARRAY ['A2', 'B1', 'C2', 'D1'], 30),
    (ARRAY ['A2', 'B1', 'C2', 'D2'], 31),
    (ARRAY ['A2', 'B1', 'C2', 'D3'], 32),
    (ARRAY ['A2', 'B1', 'C3', 'D1'], 33),
    (ARRAY ['A2', 'B1', 'C3', 'D2'], 34),
    (ARRAY ['A2', 'B1', 'C3', 'D3'], 35),
    (ARRAY ['A2', 'B2', 'C1', 'D1'], 36),
    (ARRAY ['A2', 'B2', 'C1', 'D2'], 37),
    (ARRAY ['A2', 'B2', 'C1', 'D3'], 38),
    (ARRAY ['A2', 'B2', 'C2', 'D1'], 39),
    (ARRAY ['A2', 'B2', 'C2', 'D2'], 40),
    (ARRAY ['A2', 'B2', 'C2', 'D3'], 41),
    (ARRAY ['A2', 'B2', 'C3', 'D1'], 42),
    (ARRAY ['A2', 'B2', 'C3', 'D2'], 43),
    (ARRAY ['A2', 'B2', 'C3', 'D3'], 44),
    (ARRAY ['A2', 'B3', 'C1', 'D1'], 45),
    (ARRAY ['A2', 'B3', 'C1', 'D2'], 46),
    (ARRAY ['A2', 'B3', 'C1', 'D3'], 47),
    (ARRAY ['A2', 'B3', 'C2', 'D1'], 48),
    (ARRAY ['A2', 'B3', 'C2', 'D2'], 49),
    (ARRAY ['A2', 'B3', 'C2', 'D3'], 50),
    (ARRAY ['A2', 'B3', 'C3', 'D1'], 51),
    (ARRAY ['A2', 'B3', 'C3', 'D2'], 52),
    (ARRAY ['A2', 'B3', 'C3', 'D3'], 53),
    (ARRAY ['A3', 'B1', 'C1', 'D1'], 54),
    (ARRAY ['A3', 'B1', 'C1', 'D2'], 55),
    (ARRAY ['A3', 'B1', 'C1', 'D3'], 56),
    (ARRAY ['A3', 'B1', 'C2', 'D1'], 57),
    (ARRAY ['A3', 'B1', 'C2', 'D2'], 58),
    (ARRAY ['A3', 'B1', 'C2', 'D3'], 59),
    (ARRAY ['A3', 'B1', 'C3', 'D1'], 60),
    (ARRAY ['A3', 'B1', 'C3', 'D2'], 61),
    (ARRAY ['A3', 'B1', 'C3', 'D3'], 62),
    (ARRAY ['A3', 'B2', 'C1', 'D1'], 63),
    (ARRAY ['A3', 'B2', 'C1', 'D2'], 64),
    (ARRAY ['A3', 'B2', 'C1', 'D3'], 65),
    (ARRAY ['A3', 'B2', 'C2', 'D1'], 66),
    (ARRAY ['A3', 'B2', 'C2', 'D2'], 67),
    (ARRAY ['A3', 'B2', 'C2', 'D3'], 68),
    (ARRAY ['A3', 'B2', 'C3', 'D1'], 69),
    (ARRAY ['A3', 'B2', 'C3', 'D2'], 70),
    (ARRAY ['A3', 'B2', 'C3', 'D3'], 71),
    (ARRAY ['A3', 'B3', 'C1', 'D1'], 72),
    (ARRAY ['A3', 'B3', 'C1', 'D2'], 73),
    (ARRAY ['A3', 'B3', 'C1', 'D3'], 74),
    (ARRAY ['A3', 'B3', 'C2', 'D1'], 75),
    (ARRAY ['A3', 'B3', 'C2', 'D2'], 76),
    (ARRAY ['A3', 'B3', 'C2', 'D3'], 77),
    (ARRAY ['A3', 'B3', 'C3', 'D1'], 78),
    (ARRAY ['A3', 'B3', 'C3', 'D2'], 79),
    (ARRAY ['A3', 'B3', 'C3', 'D3'], 80);

SELECT 'inlined' AS query;
SELECT path[1], path[2], path[3], path, sum(value)
FROM bind_group_by
GROUP BY GROUPING SETS (
    (path[1], path[2], path[3], path),
    (path[1], path[2], path[3]),
    (path[1], path[2]),
    (path[1]),
    ()
)
ORDER BY 1, 2, 3, 4
;

SELECT 'prepared' AS query;
PREPARE prepared_group_by (int, int, int, int, int, int, int, int, int, int, int, int) AS
    SELECT path[$1], path[$2], path[$3], path, sum(value)
    FROM bind_group_by
    GROUP BY GROUPING SETS (
        (path[$4], path[$5], path[$6], path),
        (path[$7], path[$8], path[$9]),
        (path[$10], path[$11]),
        (path[$12]),
        ()
    )
    ORDER BY 1, 2, 3, 4
;
EXECUTE prepared_group_by (1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 2, 1);

-- cleanup
DEALLOCATE prepared_group_by;
DROP TABLE bind_group_by;

Reply | Threaded
Open this post in threaded view
|

Re: Unexpected results when using GROUP BY GROUPING SETS and bind variables

Tom Lane-2
Aner Perez <[hidden email]> writes:
> [ these queries don't give the same results: ]

> SELECT path[1], path[2], path[3], path, sum(value)
> FROM bind_group_by
> GROUP BY GROUPING SETS (
>     (path[1], path[2], path[3], path),
>     (path[1], path[2], path[3]),
>     (path[1], path[2]),
>     (path[1]),
>     ()
> )
> ORDER BY 1, 2, 3, 4
> ;

> PREPARE prepared_group_by (int, int, int, int, int, int, int, int, int,
> int, int, int) AS
>     SELECT path[$1], path[$2], path[$3], path, sum(value)
>     FROM bind_group_by
>     GROUP BY GROUPING SETS (
>         (path[$4], path[$5], path[$6], path),
>         (path[$7], path[$8], path[$9]),
>         (path[$10], path[$11]),
>         (path[$12]),
>         ()
>     )
>     ORDER BY 1, 2, 3, 4
> ;
> EXECUTE prepared_group_by (1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 2, 1);

I think you're kind of shooting yourself in the foot here by writing
a query with more than one possible interpretation.  Since you have
GROUP BY items that are both the whole "path" array and specific
elements of it, it's not clear whether a targetlist reference such
as "path[1]" is meant to refer to the GROUP BY item "path[1]", or
to be a subscripting operator applied to the GROUP BY item "path".

It appears that the parser makes the first choice in your first
query, although that seems like an implementation detail that
I wouldn't care to bet on going forward.

In the second query, the parser is certainly not going to think that
"path[$1]" matches "path[$4]" or any of the other single-element
GROUP BY items, so it decides that it means subscripting the "path"
item.  Now the result will be NULL for any grouping set that doesn't
include the "path" item.

I haven't gone through the results one-by-one, but just by eyeball
I think this is sufficient to explain the discrepancies.

(I note that the SQL standard only allows GROUP BY items to be
simple column references.  Maybe that's not an arbitrary restriction
but is meant to forestall this sort of ambiguity?  Hard to be sure.)

                        regards, tom lane


Reply | Threaded
Open this post in threaded view
|

Re: Unexpected results when using GROUP BY GROUPING SETS and bind variables

Aner Perez
Thanks Tom,

I figured as much about the second query but I thought it would be safe to use the first version with the inlined indexing.
I'm not sure if you're saying that the same query without the unindexed path column in the select would be safe.  Like this:

-- Do not GROUP BY or SELECT on path by itself --
SELECT path[1], path[2], path[3], path[4], sum(value)
FROM bind_group_by
GROUP BY GROUPING SETS (
    (path[1], path[2], path[3], path[4]),
    (path[1], path[2], path[3]),
    (path[1], path[2]),
    (path[1]),
    ()
)
ORDER BY 1, 2, 3, 4;

Or if using indexed path elements in the GROUP BY is the issue and I should put the array indexing in a subselect and do the aggregation in the outer select.  Like this:

-- Safer Subselect Version --
SELECT level1, level2, level3, level4, sum(value)
FROM (
    SELECT path[1] as level1, path[2] as level2, path[3] as level3, path[4] as level4, value
    FROM bind_group_by
) AS expanded
GROUP BY GROUPING SETS (
    (level1, level2, level3, level4),
    (level1, level2, level3),
    (level1, level2),
    (level1),
    ()
)
ORDER BY 1, 2, 3, 4;

Thanks for the insight,

  - Aner

On Wed, Sep 9, 2020 at 1:13 PM Tom Lane <[hidden email]> wrote:
Aner Perez <[hidden email]> writes:
> [ these queries don't give the same results: ]

> SELECT path[1], path[2], path[3], path, sum(value)
> FROM bind_group_by
> GROUP BY GROUPING SETS (
>     (path[1], path[2], path[3], path),
>     (path[1], path[2], path[3]),
>     (path[1], path[2]),
>     (path[1]),
>     ()
> )
> ORDER BY 1, 2, 3, 4
> ;

> PREPARE prepared_group_by (int, int, int, int, int, int, int, int, int,
> int, int, int) AS
>     SELECT path[$1], path[$2], path[$3], path, sum(value)
>     FROM bind_group_by
>     GROUP BY GROUPING SETS (
>         (path[$4], path[$5], path[$6], path),
>         (path[$7], path[$8], path[$9]),
>         (path[$10], path[$11]),
>         (path[$12]),
>         ()
>     )
>     ORDER BY 1, 2, 3, 4
> ;
> EXECUTE prepared_group_by (1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 2, 1);

I think you're kind of shooting yourself in the foot here by writing
a query with more than one possible interpretation.  Since you have
GROUP BY items that are both the whole "path" array and specific
elements of it, it's not clear whether a targetlist reference such
as "path[1]" is meant to refer to the GROUP BY item "path[1]", or
to be a subscripting operator applied to the GROUP BY item "path".

It appears that the parser makes the first choice in your first
query, although that seems like an implementation detail that
I wouldn't care to bet on going forward.

In the second query, the parser is certainly not going to think that
"path[$1]" matches "path[$4]" or any of the other single-element
GROUP BY items, so it decides that it means subscripting the "path"
item.  Now the result will be NULL for any grouping set that doesn't
include the "path" item.

I haven't gone through the results one-by-one, but just by eyeball
I think this is sufficient to explain the discrepancies.

(I note that the SQL standard only allows GROUP BY items to be
simple column references.  Maybe that's not an arbitrary restriction
but is meant to forestall this sort of ambiguity?  Hard to be sure.)

                        regards, tom lane
Reply | Threaded
Open this post in threaded view
|

Re: Unexpected results when using GROUP BY GROUPING SETS and bind variables

Tom Lane-2
Aner Perez <[hidden email]> writes:
> Or if using indexed path elements in the GROUP BY is the issue and I should
> put the array indexing in a subselect and do the aggregation in the outer
> select.  Like this:

> -- Safer Subselect Version --
> SELECT level1, level2, level3, level4, sum(value)
> FROM (
>     SELECT path[1] as level1, path[2] as level2, path[3] as level3, path[4]
> as level4, value
>     FROM bind_group_by
> ) AS expanded
> GROUP BY GROUPING SETS (
>     (level1, level2, level3, level4),
>     (level1, level2, level3),
>     (level1, level2),
>     (level1),
>     ()
> )
> ORDER BY 1, 2, 3, 4;

Yeah, that one looks a lot safer from here.  There's no question about
which expressions are supposed to match what.  It should end up with
the same plan, too.

                        regards, tom lane


Reply | Threaded
Open this post in threaded view
|

Re: Unexpected results when using GROUP BY GROUPING SETS and bind variables

Aner Perez
Of course, right when I hit send I thought of another option that makes the SQL a little less readable but perhaps gets rid of the ambiguity.  Using ordinals in the GROUP BY:

SELECT path[1], path[2], path[3], path[4], sum(value)
FROM bind_group_by
GROUP BY GROUPING SETS (
    (1, 2, 3, 4),
    (1, 2, 3),
    (1, 2),
    (1),
    ()
)
ORDER BY 1, 2, 3, 4;

Since I'm generating the SQL I'm not too worried about the readability and it works with bind variables too.  I promise I'll stop thinking after this one ;-)

Thanks!

  - Aner

On Wed, Sep 9, 2020 at 2:45 PM Tom Lane <[hidden email]> wrote:
Aner Perez <[hidden email]> writes:
> Or if using indexed path elements in the GROUP BY is the issue and I should
> put the array indexing in a subselect and do the aggregation in the outer
> select.  Like this:

> -- Safer Subselect Version --
> SELECT level1, level2, level3, level4, sum(value)
> FROM (
>     SELECT path[1] as level1, path[2] as level2, path[3] as level3, path[4]
> as level4, value
>     FROM bind_group_by
> ) AS expanded
> GROUP BY GROUPING SETS (
>     (level1, level2, level3, level4),
>     (level1, level2, level3),
>     (level1, level2),
>     (level1),
>     ()
> )
> ORDER BY 1, 2, 3, 4;

Yeah, that one looks a lot safer from here.  There's no question about
which expressions are supposed to match what.  It should end up with
the same plan, too.

                        regards, tom lane