Thread: BUG #18948: Equivalent MAX() result in view and inline query yields inconsistent row count
BUG #18948: Equivalent MAX() result in view and inline query yields inconsistent row count
From
PG Bug reporting form
Date:
The following bug has been logged on the website: Bug reference: 18948 Logged by: Wenking Deng Email address: dengwenking@gmail.com PostgreSQL version: 15.0 Operating system: Docker Description: This bug demonstrates a discrepancy in PostgreSQL when comparing the result of an aggregate function used inline versus stored in a view. The queries SELECT MAX(c0) FROM t0 and SELECT c0 FROM v1 produce exactly the same value. However, when applying a WHERE (1 = NULL) condition, the inline version returns one row (with NULL), while the view version returns zero rows. Reproduce Steps: -- Setup: Create table t0 and view v1 CREATE TABLE t0 (c0 REAL); INSERT INTO t0 (c0) VALUES (1.830250668324684e+38); INSERT INTO t0 (c0) VALUES (-3.29559603270012e+38); INSERT INTO t0 (c0) VALUES (3.352134689102562e+38); CREATE VIEW v1 AS SELECT MAX(c0) AS c0 FROM t0; -- Query 1: Inline MAX() without WHERE db1=# SELECT MAX(c0) FROM t0; max --------------- 3.3521347e+38 (1 row) -- Query 2: View value (same result) db1=# SELECT c0 FROM v1; c0 --------------- 3.3521347e+38 (1 row) -- Query 3: Inline MAX() with WHERE (1 = NULL) db1=# SELECT MAX(c0) FROM t0 WHERE (1 = NULL); max ----- (1 row) -- Query 4: View value with same WHERE clause db1=# SELECT (c0) FROM v1 WHERE (1 = NULL); c0 ---- (0 rows) Expected Result: Both SELECT should return one row with NULL.
Re: BUG #18948: Equivalent MAX() result in view and inline query yields inconsistent row count
From
Tom Lane
Date:
PG Bug reporting form <noreply@postgresql.org> writes: > -- Setup: Create table t0 and view v1 > CREATE TABLE t0 (c0 REAL); > INSERT INTO t0 (c0) VALUES (1.830250668324684e+38); > INSERT INTO t0 (c0) VALUES (-3.29559603270012e+38); > INSERT INTO t0 (c0) VALUES (3.352134689102562e+38); > CREATE VIEW v1 AS SELECT MAX(c0) AS c0 FROM t0; > -- Query 3: Inline MAX() with WHERE (1 = NULL) > db1=# SELECT MAX(c0) FROM t0 WHERE (1 = NULL); > max > ----- > (1 row) > -- Query 4: View value with same WHERE clause > db1=# SELECT (c0) FROM v1 WHERE (1 = NULL); > c0 > ---- > (0 rows) > Both SELECT should return one row with NULL. No, they should not. In your query 3, the WHERE filter applies before aggregation happens, so it removes all the input rows to the MAX(). In your query 4, the WHERE filter applies after the MAX(), that is it acts on the aggregated row(s). To write an exact equivalent of query 4 without using a view or sub-select, you'd need to write the filter condition in HAVING. regards, tom lane