一个查询可从不同表中获取多个不相关的列

Question

Say I have 3 tables: table1, table2, and table3. I essentially want to combine the following 3 queries together:

1. SELECT columnA FROM table1
2. SELECT columnB FROM table2
3. SELECT columnC FROM table3

The columns may be of different size, and I want to avoid repeating data. So if columnA has 3 elements, columnB has 1 element, and columnC has 2 elements, I would get a table in which each column has the respective number of elements. Unfilled rows can be left NULL.

How do I create a single query that accomplishes this task?

The simple query SELECT DISTINCT table1.columnA, table2.columnB, table3.columnC FROM table1, table2, table3 just gives me a table of all possible combinations of all the elements within these columns.

Answer 1

UNION ALL or UNION help you. depends on if you want dupplicates or not:

SELECT columnA FROM table1
UNION ALL
SELECT columnB FROM table2
UNION ALL
SELECT columnC FROM table3

For more informations read: https://dev.mysql.com/doc/refman/8.0/en/union.html

EDIT:

You can use a second column with fix text to differentiate the records in your application:

SELECT columnA,'TABLE1' FROM table1
UNION ALL
SELECT columnB,'TABLE2' FROM table2
UNION ALL
SELECT columnC,'TABLE3' FROM table3

Answer 2

If you need only the distinct values you can use UNION

SELECT columnA , 'T1' t_name
FROM table1
UNION
SELECT columnB, 'T2'  
FROM table2
UNION
SELECT columnC , 'T3' 
FROM table3

but if you need all the values you should use UNION ALL

SELECT columnA, 'T1' t_name
FROM table1
UNION ALL
SELECT columnB, 'T2' 
FROM table2
UNION ALL
SELECT columnC , 'T3' 
FROM table3

Answer 3

If you take the union idea and add in a rownumber you can create a list of rownumbers which you can then use to left join to the three tables. For example

drop table if exists t1,t2,t3;
create table t1(id int);
create table t2(id int);
create table t3(id int);

insert into t1 values(1),(3);
insert into t2 values(10),(20),(40);
insert into t3 values(3);


select s.rownumber,t1.id t1id
        ,t2.id t2id
        ,t3.id t3id
from
(
select @r1:=@r1+1 rownumber from t1 cross join(select @r1:=0) r1
union 
select @r2:=@r2+1 rownumber from t2 cross join(select @r2:=0) r2
union 
select @r2:=@r3+1 rownumber from t3 cross join(select @r3:=0) r3
) s
left join (
select id,@r4:=@r4+1 rownumber from t1 cross join(select @r4:=0) r4
) t1 on t1.rownumber = s.rownumber
left join (
select id,@r5:=@r5+1 rownumber from t2 cross join(select @r5:=0) r5
) t2 on t2.rownumber = s.rownumber
left join (
select id,@r6:=@r6+1 rownumber from t3 cross join(select @r6:=0) r6
) t3 on t3.rownumber = s.rownumber
;

+-----------+------+------+------+
| rownumber | t1id | t2id | t3id |
+-----------+------+------+------+
|         1 |    1 |   10 |    3 |
|         2 |    3 |   20 | NULL |
|         3 | NULL |   40 | NULL |
+-----------+------+------+------+
3 rows in set (0.00 sec)

Answer 4

You can try using union

    SELECT columnA FROM table1
    union
    SELECT columnB FROM table2
    union 
    SELECT columnC FROM table3