If you execute this query
SELECT 'test-a1' AS name1, 'test-a2' AS name2
the result will be a one row-selection with two columns having these values:
test-a1, test-a2
How can I modify the above query to have a selection with several rows, e.g.
test-a1, test-a2 test-b1, test-b2 test-c1, test-c2
I know how to do this with UNION but I feel that there exists a more simple way to do it.
PS. Sorry for such a basic question, it is very hard to google it.
Answers:
Thank you for visiting the Q&A section on Magenaut. Please note that all the answers may not help you solve the issue immediately. So please treat them as advisements. If you found the post helpful (or not), leave a comment & I’ll get back to you as soon as possible.
Method 1
Values keyword can be used as below.
select * from (values ('test-a1', 'test-a2'), ('test-b1', 'test-b2'), ('test-c1', 'test-c2')) x(col1, col2)
Method 2
The following will work for SQL:
SELECT 'test-a1' AS name1, 'test-a2' AS name2 UNION ALL SELECT 'test-b1', 'test-b2' UNION ALL SELECT 'test-c1', 'test-c2'
Method 3
UNION ALL is the best bet. It’s faster than UNION and you will have mutually exclusive rows.
Method 4
Extending the answer of @openshac for oracle, as the below mentioned code works for oracle:
SELECT 'test-a1' AS name1, 'test-a2' AS name2 from dual UNION ALL SELECT 'test-b1', 'test-b2' from dual UNION ALL SELECT 'test-c1', 'test-c2' from dual
Method 5
You can use a temp table, fill it up with your results and then select from it
create table #tmpAAA (name1 varchar(10), name2 varchar(10)) insert into #tmpAAA (name1, name2) values ('test_a', 'test_b'), ('test_c', 'test_d'), ('test_e', 'test_f'), ('test_g', 'test_h'), ('test_i', 'test_j'); select * from #tmpAAA;
This will return
name1 name2 ================== test_a test_b test_c test_d test_e test_f test_g test_h test_i test_j
Method 6
I’d love to hear is anyone has a better solution. In the past I’ve used this:
Select top 3 'Hardcode' from tableWithLotsOfRows
Would you mind switching abc, with 123?
select top 3 'test-A'+convert(varchar, row_number() over (order by PrimaryKey)), 'test-B'+convert(varchar, row_number() over (order by PrimaryKey)) from tableWithLotsOfRows
that should return something like:
TestA1, Test-B1 TestA2, Test-B2 TestA3, Test-B3
Method 7
As of MySQL 8.0.19, it is possible to do
SELECT column_0 AS name1, column_1 AS name2 FROM (VALUES ROW('test-a1','test-a2'), ROW('test-b1','test-b2'), ROW('test-c1','test-c2') ) AS hardcodedNames
Which returns
name1 name2 ================== test-a1 test-a2 test-b1 test-b2 test-c1 test-c2
A note on column names
The columns of the table output from VALUES have the implicitly named columns column_0, column_1, column_2, and so on, always beginning with 0.
Documentation here: https://dev.mysql.com/doc/refman/8.0/en/values.html.
Method 8
In MySQL you could use UNION
like this:
SELECT * from (SELECT 2006 AS year UNION SELECT 2007 AS year UNION SELECT 2008 AS year UNION ) AS years
Method 9
The following code work for me in MSSQL environment:
SELECT Name1,Name2 FROM(VALUES ('test-a1', 'test-a2'), ('test-b1', 'test-b2'), ('test-c1', 'test-c2'))AS Test(Name1,Name2)
Output:
Name1 Name2 ------- ------- test-a1 test-a2 test-b1 test-b2 test-c1 test-c2
All methods was sourced from stackoverflow.com or stackexchange.com, is licensed under cc by-sa 2.5, cc by-sa 3.0 and cc by-sa 4.0