Stuck on an issue?

Lightrun Answers was designed to reduce the constant googling that comes with debugging 3rd party libraries. It collects links to all the places you might be looking at while hunting down a tough bug.

And, if you’re still stuck at the end, we’re happy to hop on a call to see how we can help out.

Data not being propagated to nested queries on where clause

See original GitHub issue

Describe the bug

The title is quite self-explanatory

column "foo.id" does not exist

🐜 This seems to be an execution error, which means that your request syntax seems okay,
    but the resulting statement cannot be executed → Probably not a pg-mem error.

*️⃣ Reconsituted failed SQL statement: SELECT (count (*) )  FROM foo   WHERE (exists ((SELECT (1)  FROM bar   WHERE (bar .foo_id = foo .id) )) )

To Reproduce

CREATE TABLE foo
(
    id    uuid NOT NULL PRIMARY KEY,
    value text NOT NULL
);

CREATE TABLE bar
(
    id uuid NOT NULL PRIMARY KEY,
    foo_id uuid NOT NULL,

    FOREIGN KEY (foo_id) REFERENCES foo(id) ON DELETE CASCADE
);

INSERT INTO foo (id, value) VALUES ('09f67c99-7c8c-40dc-8f15-13599005c657', 'foo one'), ('792f71e1-2c32-4ebe-b4c8-6cf831ce490e', 'foo two');

INSERT INTO bar (id, foo_id) VALUES ('b3381f77-8324-4778-9bd5-74d90fe14553', '792f71e1-2c32-4ebe-b4c8-6cf831ce490e');

SELECT COUNT(*) FROM foo WHERE EXISTS (
    SELECT 1 FROM bar WHERE bar.foo_id = foo.id
)

On postgres:

postgres=*# CREATE TABLE foo
postgres-*# (
postgres(*#     id    uuid NOT NULL PRIMARY KEY,
postgres(*#     value text NOT NULL
postgres(*# );
CREATE TABLE
postgres=*# 
postgres=*# CREATE TABLE bar
postgres-*# (
postgres(*#     id uuid NOT NULL PRIMARY KEY,
postgres(*#     foo_id uuid NOT NULL,
postgres(*# 
postgres(*#     FOREIGN KEY (foo_id) REFERENCES foo(id) ON DELETE CASCADE
postgres(*# );
CREATE TABLE
postgres=*# 
postgres=*# INSERT INTO foo (id, value) VALUES ('09f67c99-7c8c-40dc-8f15-13599005c657', 'foo one'), ('792f71e1-2c32-4ebe-b4c8-6cf831ce490e', 'foo two');
INSERT 0 2
postgres=*# 
postgres=*# INSERT INTO bar (id, foo_id) VALUES ('b3381f77-8324-4778-9bd5-74d90fe14553', '792f71e1-2c32-4ebe-b4c8-6cf831ce490e');
INSERT 0 1
postgres=*# 
postgres=*# SELECT COUNT(*) FROM foo WHERE EXISTS (
postgres(*#     SELECT 1 FROM bar WHERE bar.foo_id = foo.id
postgres(*# );
 count 
-------
     1
(1 row)

pg-mem version

2.6.3

Issue Analytics

State:
Created a year ago
Comments:5

Top GitHub Comments

1reaction

Fryunicommented, Nov 3, 2022

There is no way to get automatically get the optimized query that I know of.

I called EXPLAIN (ANALYZE, BUFFERS, VERBOSE, COSTS, TIMING, FORMAT JSON) passed it through a plan visualizer (this one). From that I saw which joins were performed to get the same result and rewrote to query with them making sure that the resulting plan was the same or equivalent.

0reactions

oseibonsucommented, Nov 3, 2022

@Fryuni I am somewhat new to Postgres is there a command to have it output the optimized planned query?