1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
|
--source include/elide_costs.inc
--echo #
--echo # Bug#35842412 WHERE NOT IN with subquery is much slower on 8.1 than 5.7
--echo #
CREATE TABLE t1( b_id INTEGER NOT NULL PRIMARY KEY AUTO_INCREMENT,
val VARCHAR(100) NOT NULL, val_n VARCHAR(100),
KEY ix_val (val, b_id), KEY ix_val_n (val_n, b_id));
CREATE TABLE t2( a_id INTEGER NOT NULL PRIMARY KEY AUTO_INCREMENT,
val VARCHAR(100) NOT NULL, val_n VARCHAR(100),
UNIQUE KEY ix_val (val), UNIQUE KEY ix_val_n (val_n));
INSERT INTO t1(val, val_n)
WITH RECURSIVE DataSource(num, c1, c2) AS (
SELECT 1, lpad(1,25,'0'), lpad(1,25,'0')
UNION ALL
SELECT num + 1, lpad(mod(num+1,30),25,'0'),
lpad(ELT(num%400 != 0, mod(num+1,30)), 25, '0')
FROM DataSource
WHERE num < 1000
)
SELECT c1, c2 from DataSource;
INSERT INTO t2(val, val_n)
WITH RECURSIVE DataSource(num, c1, c2) AS (
SELECT 1, lpad(1,25,'0'), lpad(1,25,'0')
UNION ALL
SELECT num + 1, lpad(num+1,25,'0'), lpad(ELT(num%13 != 0, num+1), 25, '0')
FROM DataSource
WHERE num < 30
)
SELECT c1, c2 from DataSource;
ANALYZE TABLE t1, t2;
--echo # Should use group skip scan on t1
let query = SELECT DISTINCT val FROM t1 WHERE EXISTS
(SELECT * FROM t2 WHERE t2.val = t1.val);
--replace_regex $elide_costs
eval EXPLAIN FORMAT=TREE $query;
eval $query;
let query = SELECT val FROM t1 WHERE EXISTS
(SELECT * FROM t2 WHERE t2.val = t1.val) GROUP BY val;
--replace_regex $elide_costs
eval EXPLAIN FORMAT=TREE $query;
eval $query;
let query = SELECT DISTINCT val FROM t1 WHERE val IN (SELECT val FROM t2);
--replace_regex $elide_costs
eval EXPLAIN FORMAT=TREE $query;
eval $query;
let query = SELECT DISTINCT val FROM t1 WHERE val IN
(SELECT val FROM t2 WHERE t2.val = t1.val);
--replace_regex $elide_costs
eval EXPLAIN FORMAT=TREE $query;
eval $query;
let query = SELECT val FROM t1 WHERE val IN (SELECT val FROM t2 ) GROUP BY val;
--replace_regex $elide_costs
eval EXPLAIN FORMAT=TREE $query;
eval $query;
let query = SELECT val FROM t1 WHERE val IN
(SELECT val FROM t2 WHERE t2.val = t1.val ) GROUP BY val;
--replace_regex $elide_costs
eval EXPLAIN FORMAT=TREE $query;
eval $query;
let query = SELECT DISTINCT val FROM t1 WHERE NOT EXISTS
(SELECT * FROM t2 WHERE val = 'asd');
--replace_regex $elide_costs
eval EXPLAIN FORMAT=TREE $query;
eval $query;
let query = SELECT val FROM t1 WHERE NOT EXISTS
(SELECT * FROM t2 WHERE val = 'asd' ) GROUP BY val;
--replace_regex $elide_costs
eval EXPLAIN FORMAT=TREE $query;
eval $query;
--echo # Should use group skip scan on t1
let query = SELECT DISTINCT val FROM t1 WHERE val NOT IN (SELECT val FROM t2);
--replace_regex $elide_costs
eval EXPLAIN FORMAT=TREE $query;
eval $query;
let query = SELECT DISTINCT val FROM t1 WHERE val NOT IN
(SELECT val FROM t2 WHERE t2.val = t1.val);
--replace_regex $elide_costs
eval EXPLAIN FORMAT=TREE $query;
eval $query;
let query = SELECT val FROM t1 WHERE val NOT IN
(SELECT val FROM t2 ) GROUP BY val;
--replace_regex $elide_costs
eval EXPLAIN FORMAT=TREE $query;
eval $query;
let query = SELECT val FROM t1 WHERE val NOT IN
(SELECT val FROM t2 WHERE t2.val = t1.val ) GROUP BY val;
--replace_regex $elide_costs
eval EXPLAIN FORMAT=TREE $query;
eval $query;
let query = SELECT DISTINCT val FROM t1 WHERE NOT EXISTS
(SELECT * FROM t2 WHERE t2.val = t1.val);
--replace_regex $elide_costs
eval EXPLAIN FORMAT=TREE $query;
eval $query;
let query = SELECT val FROM t1 WHERE NOT EXISTS
(SELECT * FROM t2 WHERE t2.val = t1.val) GROUP BY val;
--replace_regex $elide_costs
eval EXPLAIN FORMAT=TREE $query;
eval $query;
--echo # Should use group skip scan on t1
let query = SELECT DISTINCT val FROM t1 WHERE EXISTS (SELECT * FROM t2);
--replace_regex $elide_costs
eval EXPLAIN FORMAT=TREE $query;
eval $query;
let query = SELECT val FROM t1 WHERE EXISTS (SELECT * FROM t2) GROUP BY val;
--replace_regex $elide_costs
eval EXPLAIN FORMAT=TREE $query;
eval $query;
--echo # Uses group skip scan with aggregate functions since "t2" is a const
--echo # table. Hypergraph optimizer does not support const table optimization.
--echo # So the plan chosen is different.
let query = SELECT val, MIN(b_id), MAX(b_id) FROM t1 WHERE NOT EXISTS
(SELECT * FROM t2 WHERE val = 'asd') GROUP BY val;
--replace_regex $elide_costs
eval EXPLAIN FORMAT=TREE $query;
eval $query;
--echo # Should not use group skip scan on t1. Group skip scan is not supported
--echo # on multi-table queries with aggregate functions.
let query = SELECT val, MIN(b_id), MAX(b_id) FROM t1 WHERE val
NOT IN (SELECT val FROM t2) GROUP BY val;
--replace_regex $elide_costs
eval EXPLAIN FORMAT=TREE $query;
eval $query;
let query = SELECT val, MIN(b_id), MAX(b_id) FROM t1 WHERE EXISTS
(SELECT * FROM t2) GROUP BY val;
--replace_regex $elide_costs
eval EXPLAIN FORMAT=TREE $query;
eval $query;
--echo # Cover nullable column scenarios
--echo # Should use group skip scan on t1
let query = SELECT DISTINCT val_n FROM t1 WHERE EXISTS
(SELECT * FROM t2 WHERE t2.val_n = t1.val_n);
--replace_regex $elide_costs
eval EXPLAIN FORMAT=TREE $query;
eval $query;
let query = SELECT DISTINCT val_n FROM t1 WHERE val_n IN (SELECT val_n FROM t2);
--replace_regex $elide_costs
eval EXPLAIN FORMAT=TREE $query;
eval $query;
let query = SELECT DISTINCT val_n FROM t1 WHERE NOT EXISTS
(SELECT * FROM t2 WHERE val_n = 'asd');
--replace_regex $elide_costs
eval EXPLAIN FORMAT=TREE $query;
eval $query;
--echo # Should use group skip scan on t1
let query = SELECT DISTINCT val_n FROM t1 WHERE val_n NOT IN
(SELECT val_n FROM t2);
--replace_regex $elide_costs
eval EXPLAIN FORMAT=TREE $query;
eval $query;
let query = SELECT DISTINCT val_n FROM t1 WHERE val_n NOT IN
(SELECT val_n FROM t2 WHERE t2.val_n = t1.val_n);
--replace_regex $elide_costs
eval EXPLAIN FORMAT=TREE $query;
eval $query;
let query = SELECT DISTINCT val_n FROM t1 WHERE NOT EXISTS
(SELECT * FROM t2 WHERE t2.val_n = t1.val_n);
--replace_regex $elide_costs
eval EXPLAIN FORMAT=TREE $query;
eval $query;
let query = SELECT val_n FROM t1 WHERE NOT EXISTS
(SELECT * FROM t2 WHERE t2.val_n = t1.val_n) GROUP BY val_n;
--replace_regex $elide_costs
eval EXPLAIN FORMAT=TREE $query;
eval $query;
--echo # Should use group skip scan on t1
let query = SELECT DISTINCT val_n FROM t1 WHERE EXISTS (SELECT * FROM t2);
--replace_regex $elide_costs
eval EXPLAIN FORMAT=TREE $query;
eval $query;
--echo # Uses group skip scan with aggregate functions since "t2" is a const
--echo # table. Hypergraph optimizer does not support const table optimization.
--echo # So the plan chosen is different.
let query = SELECT val_n, MIN(b_id), MAX(b_id) FROM t1 WHERE NOT EXISTS
(SELECT * FROM t2 WHERE val_n = 'asd') GROUP BY val_n;
--replace_regex $elide_costs
eval EXPLAIN FORMAT=TREE $query;
eval $query;
--echo # Should not use group skip scan on t1. Group skip scan is not supported
--echo # on multi-table queries with aggregate functions.
let query = SELECT val_n, MIN(b_id), MAX(b_id) FROM t1 WHERE val_n
NOT IN (SELECT val_n FROM t2) GROUP BY val_n;
--replace_regex $elide_costs
eval EXPLAIN FORMAT=TREE $query;
eval $query;
let query = SELECT val_n, MIN(b_id), MAX(b_id) FROM t1 WHERE EXISTS
(SELECT * FROM t2) GROUP BY val_n;
--replace_regex $elide_costs
eval EXPLAIN FORMAT=TREE $query;
eval $query;
DROP TABLE t1, t2;
|
