Commit 55499d2b authored by unknown's avatar unknown

Fix for bug #32202: ORDER BY not working with GROUP BY

The bug is a regression introduced by the fix for bug30596. The problem
was that in cases when groups in GROUP BY correspond to only one row,
and there is ORDER BY, the GROUP BY was removed and the ORDER BY
rewritten to ORDER BY <group_by_columns> without checking if the
columns in GROUP BY and ORDER BY are compatible. This led to
incorrect ordering of the result set as it was sorted using the
GROUP BY columns. Additionaly, the code discarded ASC/DESC modifiers
from ORDER BY even if its columns were compatible with the GROUP BY
ones.

This patch fixes the regression by checking if ORDER BY columns form a
prefix of the GROUP BY ones, and rewriting ORDER BY only in that case,
preserving the ASC/DESC modifiers. That check is sufficient, since the
GROUP BY columns contain a unique index.


mysql-test/r/group_by.result:
  Added a test case for bug #32202.
mysql-test/t/group_by.test:
  Added a test case for bug #32202.
sql/sql_select.cc:
  In cases when groups in GROUP BY correspond to only one row and there
  is ORDER BY, rewrite the query to ORDER BY <group_by_columns> only if
  the columns in ORDER BY and GROUP BY are compatible, i.e. either one
  forms a prefix for another.
parent 4713575c
......@@ -1113,3 +1113,68 @@ c b
3 1
3 2
DROP TABLE t1;
CREATE TABLE t1(
id INT AUTO_INCREMENT PRIMARY KEY,
c1 INT NOT NULL,
c2 INT NOT NULL,
UNIQUE KEY (c2,c1));
INSERT INTO t1(c1,c2) VALUES (5,1), (4,1), (3,5), (2,3), (1,3);
SELECT * FROM t1 ORDER BY c1;
id c1 c2
5 1 3
4 2 3
3 3 5
2 4 1
1 5 1
SELECT * FROM t1 GROUP BY id ORDER BY c1;
id c1 c2
5 1 3
4 2 3
3 3 5
2 4 1
1 5 1
SELECT * FROM t1 GROUP BY id ORDER BY id DESC;
id c1 c2
5 1 3
4 2 3
3 3 5
2 4 1
1 5 1
SELECT * FROM t1 GROUP BY c2 ,c1, id ORDER BY c2, c1;
id c1 c2
2 4 1
1 5 1
5 1 3
4 2 3
3 3 5
SELECT * FROM t1 GROUP BY c2, c1, id ORDER BY c2 DESC, c1;
id c1 c2
3 3 5
5 1 3
4 2 3
2 4 1
1 5 1
SELECT * FROM t1 GROUP BY c2, c1, id ORDER BY c2 DESC, c1 DESC;
id c1 c2
3 3 5
4 2 3
5 1 3
1 5 1
2 4 1
SELECT * FROM t1 GROUP BY c2 ORDER BY c2, c1;
id c1 c2
1 5 1
4 2 3
3 3 5
SELECT * FROM t1 GROUP BY c2 ORDER BY c2 DESC, c1;
id c1 c2
3 3 5
4 2 3
1 5 1
SELECT * FROM t1 GROUP BY c2 ORDER BY c2 DESC, c1 DESC;
id c1 c2
3 3 5
4 2 3
1 5 1
DROP TABLE t1;
End of 5.0 tests
......@@ -815,3 +815,38 @@ EXPLAIN SELECT c,b FROM t1 GROUP BY c,b;
SELECT c,b FROM t1 GROUP BY c,b;
DROP TABLE t1;
#
# Bug #32202: ORDER BY not working with GROUP BY
#
CREATE TABLE t1(
id INT AUTO_INCREMENT PRIMARY KEY,
c1 INT NOT NULL,
c2 INT NOT NULL,
UNIQUE KEY (c2,c1));
INSERT INTO t1(c1,c2) VALUES (5,1), (4,1), (3,5), (2,3), (1,3);
# Show that the test cases from the bug report pass
SELECT * FROM t1 ORDER BY c1;
SELECT * FROM t1 GROUP BY id ORDER BY c1;
# Show that DESC is handled correctly
SELECT * FROM t1 GROUP BY id ORDER BY id DESC;
# Show that results are correctly ordered when ORDER BY fields
# are a subset of GROUP BY ones
SELECT * FROM t1 GROUP BY c2 ,c1, id ORDER BY c2, c1;
SELECT * FROM t1 GROUP BY c2, c1, id ORDER BY c2 DESC, c1;
SELECT * FROM t1 GROUP BY c2, c1, id ORDER BY c2 DESC, c1 DESC;
# Show that results are correctly ordered when GROUP BY fields
# are a subset of ORDER BY ones
SELECT * FROM t1 GROUP BY c2 ORDER BY c2, c1;
SELECT * FROM t1 GROUP BY c2 ORDER BY c2 DESC, c1;
SELECT * FROM t1 GROUP BY c2 ORDER BY c2 DESC, c1 DESC;
DROP TABLE t1;
--echo End of 5.0 tests
......@@ -1057,10 +1057,19 @@ JOIN::optimize()
We have found that grouping can be removed since groups correspond to
only one row anyway, but we still have to guarantee correct result
order. The line below effectively rewrites the query from GROUP BY
<fields> to ORDER BY <fields>. One exception is if skip_sort_order is
set (see above), then we can simply skip GROUP BY.
<fields> to ORDER BY <fields>. There are two exceptions:
- if skip_sort_order is set (see above), then we can simply skip
GROUP BY;
- we can only rewrite ORDER BY if the ORDER BY fields are 'compatible'
with the GROUP BY ones, i.e. either one is a prefix of another.
We only check if the ORDER BY is a prefix of GROUP BY. In this case
test_if_subpart() copies the ASC/DESC attributes from the original
ORDER BY fields.
If GROUP BY is a prefix of ORDER BY, then it is safe to leave
'order' as is.
*/
order= skip_sort_order ? 0 : group_list;
if (!order || test_if_subpart(group_list, order))
order= skip_sort_order ? 0 : group_list;
group_list= 0;
group= 0;
}
......
Markdown is supported
0%
or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment