~drizzle-trunk/drizzle/development

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
drop table if exists t1;
show variables like 'version';
Variable_name	Value
version	####.##.####
select version();
version()
####.##.####
select database();
database()
test
CREATE TABLE `t1` (
`col_bigint` bigint,
`col_text` text,
`col_char` char (1),
`col_enum` enum ('a','b','c','d','e','f','g','h','i','j','k','l','m','n','o','p','q','r','s','t','u','v','w','x','y','z'),
`col_int` int,
`col_char_not_null` char (1) not null,
`col_int_not_null_key` int not null,
`col_text_not_null` text not null,
`col_enum_not_null_key` enum ('a','b','c','d','e','f','g','h','i','j','k','l','m','n','o','p','q','r','s','t','u','v','w','x','y','z') not null,
`col_int_key` int,
`col_char_key` char (1),
`col_enum_not_null` enum ('a','b','c','d','e','f','g','h','i','j','k','l','m','n','o','p','q','r','s','t','u','v','w','x','y','z') not null,
`col_text_not_null_key` text not null,
pk integer auto_increment,
`col_bigint_key` bigint,
`col_int_not_null` int not null,
`col_bigint_not_null` bigint not null,
`col_bigint_not_null_key` bigint not null,
`col_char_not_null_key` char (1) not null,
`col_text_key` text,
`col_enum_key` enum ('a','b','c','d','e','f','g','h','i','j','k','l','m','n','o','p','q','r','s','t','u','v','w','x','y','z'),
/*Indices*/
key (`col_int_not_null_key` ),
key (`col_enum_not_null_key` ),
key (`col_int_key` ),
key (`col_char_key` ),
key (`col_text_not_null_key`  (255)),
primary key (pk),
key (`col_bigint_key` ),
key (`col_bigint_not_null_key` ),
key (`col_char_not_null_key` ),
key (`col_text_key`  (255)),
key (`col_enum_key` )) ENGINE=innodb;
set AUTOCOMMIT=OFF;
select SQL_BIG_RESULT table_schema, 
table_name, 
CASE WHEN table_type = 'STANDARD' THEN 'table' 
WHEN table_type = 'FUNCTION' then 'function' 
ELSE 'misc' END, 
column_name,
CASE WHEN IS_USED_IN_PRIMARY = 'YES' THEN 'primary' 
WHEN IS_INDEXED = 'YES' THEN 'indexed' 
ELSE 'indexed' END 
FROM data_dictionary.tables INNER JOIN
data_dictionary.columns USING(table_schema, table_name) ORDER BY table_schema, table_name limit 20;
table_schema	table_name	CASE WHEN table_type = 'STANDARD' THEN 'table' 
WHEN table_type = 'FUNCTION' then 'function' 
ELSE 'misc' END	column_name	CASE WHEN IS_USED_IN_PRIMARY = 'YES' THEN 'primary' 
WHEN IS_INDEXED = 'YES' THEN 'indexed' 
ELSE 'indexed' END 

DATA_DICTIONARY	CATALOGS	misc	CATALOG_NAME	indexed
DATA_DICTIONARY	CATALOGS	misc	CATALOG_CREATION_TIME	indexed
DATA_DICTIONARY	CATALOGS	misc	CATALOG_UPDATE_TIME	indexed
DATA_DICTIONARY	CATALOGS	misc	CATALOG_UUID	indexed
DATA_DICTIONARY	CATALOGS	misc	CATALOG_VERSION	indexed
DATA_DICTIONARY	CATALOG_CACHE	misc	CATALOG_NAME	indexed
DATA_DICTIONARY	CHARACTER_SETS	misc	CHARACTER_SET_NAME	indexed
DATA_DICTIONARY	CHARACTER_SETS	misc	DEFAULT_COLLATE_NAME	indexed
DATA_DICTIONARY	CHARACTER_SETS	misc	DESCRIPTION	indexed
DATA_DICTIONARY	CHARACTER_SETS	misc	MAXLEN	indexed
DATA_DICTIONARY	COLLATIONS	misc	CHARACTER_SET_NAME	indexed
DATA_DICTIONARY	COLLATIONS	misc	COLLATION_NAME	indexed
DATA_DICTIONARY	COLLATIONS	misc	DESCRIPTION	indexed
DATA_DICTIONARY	COLLATIONS	misc	ID	indexed
DATA_DICTIONARY	COLLATIONS	misc	IS_DEFAULT	indexed
DATA_DICTIONARY	COLLATIONS	misc	IS_COMPILED	indexed
DATA_DICTIONARY	COLLATIONS	misc	SORTLEN	indexed
DATA_DICTIONARY	COLUMNS	misc	TABLE_SCHEMA	indexed
DATA_DICTIONARY	COLUMNS	misc	TABLE_NAME	indexed
DATA_DICTIONARY	COLUMNS	misc	COLUMN_NAME	indexed
CREATE SCHEMA randgen_query_test;
USE randgen_query_test;
CREATE TABLE t1 (a INT);
CREATE TABLE t2 LIKE t1;
CREATE TABLE t3 LIKE t1;
CREATE TABLE t4 (b CHAR(500));
CREATE TABLE t5 LIKE t4;
select SQL_BIG_RESULT table_schema,
table_name,
CASE WHEN table_type = 'STANDARD' THEN 'table'
            WHEN table_type = 'FUNCTION' then 'function'
       ELSE 'misc' END,
column_name,
CASE WHEN IS_USED_IN_PRIMARY = 'YES' THEN 'primary'
             WHEN IS_INDEXED = 'YES' THEN 'indexed'
       ELSE 'indexed' END
FROM data_dictionary.tables INNER JOIN
data_dictionary.columns USING(table_schema, table_name) WHERE table_schema NOT IN ('DATA_DICTIONARY','INFORMATION_SCHEMA')
ORDER BY table_schema, table_name limit 20;
table_schema	table_name	CASE WHEN table_type = 'STANDARD' THEN 'table'
            WHEN table_type = 'FUNCTION' then 'function'
       ELSE 'misc' END	column_name	CASE WHEN IS_USED_IN_PRIMARY = 'YES' THEN 'primary'
             WHEN IS_INDEXED = 'YES' THEN 'indexed'
       ELSE 'indexed' END

randgen_query_test	t1	misc	a	indexed
randgen_query_test	t2	misc	a	indexed
randgen_query_test	t3	misc	a	indexed
randgen_query_test	t4	misc	b	indexed
randgen_query_test	t5	misc	b	indexed
test	t1	misc	col_bigint	indexed
test	t1	misc	col_text	indexed
test	t1	misc	col_char	indexed
test	t1	misc	col_enum	indexed
test	t1	misc	col_int	indexed
test	t1	misc	col_char_not_null	indexed
test	t1	misc	col_int_not_null_key	indexed
test	t1	misc	col_text_not_null	indexed
test	t1	misc	col_enum_not_null_key	indexed
test	t1	misc	col_int_key	indexed
test	t1	misc	col_char_key	indexed
test	t1	misc	col_enum_not_null	indexed
test	t1	misc	col_text_not_null_key	indexed
test	t1	misc	pk	primary
test	t1	misc	col_bigint_key	indexed
DROP SCHEMA randgen_query_test;
USE test;
select AVG(`col_int_key`) + AVG(`col_int`) AS average1,
(SUM(`col_int_key`) + SUM(`col_int`)) / COUNT(*) AS average2,
COUNT(*) AS count FROM t1;
average1	average2	count
NULL	NULL	0
select collation_name,character_set_name FROM data_dictionary.collations;
collation_name	character_set_name
binary	binary
utf8_bin	utf8_general_ci
utf8_czech_ci	utf8_general_ci
utf8_danish_ci	utf8_general_ci
utf8_esperanto_ci	utf8_general_ci
utf8_estonian_ci	utf8_general_ci
utf8_general_ci	utf8_general_ci
utf8_hungarian_ci	utf8_general_ci
utf8_icelandic_ci	utf8_general_ci
utf8_latvian_ci	utf8_general_ci
utf8_lithuanian_ci	utf8_general_ci
utf8_persian_ci	utf8_general_ci
utf8_polish_ci	utf8_general_ci
utf8_roman_ci	utf8_general_ci
utf8_romanian_ci	utf8_general_ci
utf8_sinhala_ci	utf8_general_ci
utf8_slovak_ci	utf8_general_ci
utf8_slovenian_ci	utf8_general_ci
utf8_spanish2_ci	utf8_general_ci
utf8_spanish_ci	utf8_general_ci
utf8_swedish_ci	utf8_general_ci
utf8_turkish_ci	utf8_general_ci
utf8_unicode_ci	utf8_general_ci
select COUNT(*) from data_dictionary.tables;
COUNT(*)
####.##.####
COMMIT;
DROP TABLE t1;