回答編集履歴
3
ついき
test
CHANGED
|
@@ -117,3 +117,79 @@
|
|
|
117
117
|
これをループさせて羅列する処理をprocedureで書くわけです
|
|
118
118
|
|
|
119
119
|
興味があるようならprocedureの書き方も例示します
|
|
120
|
+
|
|
121
|
+
|
|
122
|
+
|
|
123
|
+
# procedure
|
|
124
|
+
|
|
125
|
+
横方向への羅列するprocedure
|
|
126
|
+
|
|
127
|
+
|
|
128
|
+
|
|
129
|
+
```SQL
|
|
130
|
+
|
|
131
|
+
drop procedure if exists proc_yoko_hyoji;
|
|
132
|
+
|
|
133
|
+
delimiter //
|
|
134
|
+
|
|
135
|
+
create procedure proc_yoko_hyoji()
|
|
136
|
+
|
|
137
|
+
begin
|
|
138
|
+
|
|
139
|
+
declare cnt int;
|
|
140
|
+
|
|
141
|
+
declare len int;
|
|
142
|
+
|
|
143
|
+
select count(*) into len from tbl group by mid order by len desc limit 1;
|
|
144
|
+
|
|
145
|
+
set @sql="select mid";
|
|
146
|
+
|
|
147
|
+
set cnt=1;
|
|
148
|
+
|
|
149
|
+
while cnt<=len DO
|
|
150
|
+
|
|
151
|
+
set @sql=concat(@sql
|
|
152
|
+
|
|
153
|
+
,",group_concat(case rank when ",cnt," then sid end) as sid",cnt
|
|
154
|
+
|
|
155
|
+
,",group_concat(case rank when ",cnt," then snote end) as snote",cnt
|
|
156
|
+
|
|
157
|
+
);
|
|
158
|
+
|
|
159
|
+
set cnt=cnt+1;
|
|
160
|
+
|
|
161
|
+
end while;
|
|
162
|
+
|
|
163
|
+
set @sql=concat(@sql
|
|
164
|
+
|
|
165
|
+
," from ( "
|
|
166
|
+
|
|
167
|
+
,"select mid,sid,snote,(select count(*)+1 from tbl as t2 where t1.mid=t2.mid and t1.sid>t2.sid) as rank "
|
|
168
|
+
|
|
169
|
+
,"from tbl as t1 "
|
|
170
|
+
|
|
171
|
+
,") as sub "
|
|
172
|
+
|
|
173
|
+
,"group by mid "
|
|
174
|
+
|
|
175
|
+
);
|
|
176
|
+
|
|
177
|
+
prepare stmt from @sql;
|
|
178
|
+
|
|
179
|
+
execute stmt;
|
|
180
|
+
|
|
181
|
+
end
|
|
182
|
+
|
|
183
|
+
//
|
|
184
|
+
|
|
185
|
+
delimiter ;
|
|
186
|
+
|
|
187
|
+
```
|
|
188
|
+
|
|
189
|
+
※上記を一度つくっておけばあとは、データ更新などがあっても以下を実行するだけです
|
|
190
|
+
|
|
191
|
+
```SQL
|
|
192
|
+
|
|
193
|
+
call proc_yoko_hyoji;
|
|
194
|
+
|
|
195
|
+
```
|
2
調整
test
CHANGED
|
@@ -76,21 +76,21 @@
|
|
|
76
76
|
|
|
77
77
|
select mid
|
|
78
78
|
|
|
79
|
-
,group_concat(case when
|
|
79
|
+
,group_concat(case rank when 1 then sid end) as sid1
|
|
80
80
|
|
|
81
|
-
,group_concat(case when
|
|
81
|
+
,group_concat(case rank when 1 then snote end) as snote1
|
|
82
82
|
|
|
83
|
-
,group_concat(case when
|
|
83
|
+
,group_concat(case rank when 2 then sid end) as sid2
|
|
84
84
|
|
|
85
|
-
,group_concat(case when
|
|
85
|
+
,group_concat(case rank when 2 then snote end) as snote2
|
|
86
86
|
|
|
87
|
-
,group_concat(case when
|
|
87
|
+
,group_concat(case rank when 3 then sid end) as sid3
|
|
88
88
|
|
|
89
|
-
,group_concat(case when
|
|
89
|
+
,group_concat(case rank when 3 then snote end) as snote3
|
|
90
90
|
|
|
91
|
-
,group_concat(case when
|
|
91
|
+
,group_concat(case rank when 4 then sid end) as sid4
|
|
92
92
|
|
|
93
|
-
,group_concat(case when
|
|
93
|
+
,group_concat(case rank when 4 then snote end) as snote4
|
|
94
94
|
|
|
95
95
|
from(
|
|
96
96
|
|
|
@@ -101,6 +101,8 @@
|
|
|
101
101
|
) as b
|
|
102
102
|
|
|
103
103
|
group by mid
|
|
104
|
+
|
|
105
|
+
|
|
104
106
|
|
|
105
107
|
```
|
|
106
108
|
|
1
sample
test
CHANGED
|
@@ -27,3 +27,91 @@
|
|
|
27
27
|
```
|
|
28
28
|
|
|
29
29
|
のようなとり方をするとかで対応してみてください
|
|
30
|
+
|
|
31
|
+
|
|
32
|
+
|
|
33
|
+
# sample
|
|
34
|
+
|
|
35
|
+
SQLでやる簡単なサンプルを付けておきます
|
|
36
|
+
|
|
37
|
+
|
|
38
|
+
|
|
39
|
+
- 元データ
|
|
40
|
+
|
|
41
|
+
|
|
42
|
+
|
|
43
|
+
```SQL
|
|
44
|
+
|
|
45
|
+
create table tbl(mid int,sid int unique,snote varchar(20));
|
|
46
|
+
|
|
47
|
+
insert into tbl values
|
|
48
|
+
|
|
49
|
+
(1,1,'a'),
|
|
50
|
+
|
|
51
|
+
(1,2,'b'),
|
|
52
|
+
|
|
53
|
+
(1,3,'c'),
|
|
54
|
+
|
|
55
|
+
(1,4,'d'),
|
|
56
|
+
|
|
57
|
+
(2,5,'e'),
|
|
58
|
+
|
|
59
|
+
(2,6,'f'),
|
|
60
|
+
|
|
61
|
+
(2,7,'g'),
|
|
62
|
+
|
|
63
|
+
(3,8,'h'),
|
|
64
|
+
|
|
65
|
+
(4,9,'i'),
|
|
66
|
+
|
|
67
|
+
(4,10,'k');
|
|
68
|
+
|
|
69
|
+
```
|
|
70
|
+
|
|
71
|
+
|
|
72
|
+
|
|
73
|
+
- 検索
|
|
74
|
+
|
|
75
|
+
```SQL
|
|
76
|
+
|
|
77
|
+
select mid
|
|
78
|
+
|
|
79
|
+
,group_concat(case when rank=1 then sid end) as sid1
|
|
80
|
+
|
|
81
|
+
,group_concat(case when rank=1 then snote end) as snote1
|
|
82
|
+
|
|
83
|
+
,group_concat(case when rank=2 then sid end) as sid2
|
|
84
|
+
|
|
85
|
+
,group_concat(case when rank=2 then snote end) as snote2
|
|
86
|
+
|
|
87
|
+
,group_concat(case when rank=3 then sid end) as sid3
|
|
88
|
+
|
|
89
|
+
,group_concat(case when rank=3 then snote end) as snote3
|
|
90
|
+
|
|
91
|
+
,group_concat(case when rank=4 then sid end) as sid4
|
|
92
|
+
|
|
93
|
+
,group_concat(case when rank=4 then snote end) as snote4
|
|
94
|
+
|
|
95
|
+
from(
|
|
96
|
+
|
|
97
|
+
select mid,sid,snote,(select count(*)+1 from tbl as t2 where t1.mid=t2.mid and t1.sid>t2.sid) as rank
|
|
98
|
+
|
|
99
|
+
from tbl t1
|
|
100
|
+
|
|
101
|
+
) as b
|
|
102
|
+
|
|
103
|
+
group by mid
|
|
104
|
+
|
|
105
|
+
```
|
|
106
|
+
|
|
107
|
+
|
|
108
|
+
|
|
109
|
+
上記元データはmid=1のデータが4つあると目視できるため4列書いていますが
|
|
110
|
+
|
|
111
|
+
実際にはいくつあるかは状況によって違うでしょう
|
|
112
|
+
|
|
113
|
+
原則としてgroup_concatの部分は手書きでしかSQLは指定できないので、
|
|
114
|
+
|
|
115
|
+
これをループさせて羅列する処理をprocedureで書くわけです
|
|
116
|
+
|
|
117
|
+
興味があるようならprocedureの書き方も例示します
|