mysql merge rows with no unique rows - mysql

Question

I need help in merging rowns into column in sql:
Table:
|TopologyType | TopologyName | Name |
| 1 | oneway | client1 |
| 1 | oneway | client2 |
| 2 | proxy | client1 |
| 2 | proxy | proxy1 |
| 2 | proxy | client2 |
Output:
| 1 | oneway | client1 | NULL | client2 |
| 2 | proxy | client1 | proxy | client2 |

Answer 1

In case you don't need all of the output fields to be in columns you could use GROUP_CONCAT
Example query
SELECT `t`.`TopologyType`,`t`.`TopologyName`,GROUP_CONCAT(DISTINCT(`tt`.`Name`)) as `name` from `test` `t`
left join `test` `tt`
ON `t`.`TopologyType`=`tt`.`TopologyType`
GROUP BY `t`.`TopologyName`;
Output
TopologyType TopologyName name
1 oneway client1,client2
2 proxy client1,proxy1,client2
view this example in SQL fiddle

Answer 2

Oof, you're a bit ropey on detail so don't expect this answer to work without some adjustment. Let me present you some sample data:
Name, Class, Score
john, math, 70
james, math, 75
john, english, 48
james, english, 69
Desired output:
name, mathscore, englishscore
john, 70, 48
james, 75, 69
Pivoting query for mysql (doesn't support PIVOT command)
SELECT
name,
MAX(CASE WHEN class = 'math' THEN score END) as mathscore
MAX(CASE WHEN class = 'english' THEN score END) as englishscore
FROM
grades
GROUP BY
name
If you want to know more about how it works, then run it without the GROUP BY and without the MAX functions:
SELECT
name,
CASE WHEN class = 'math' THEN score END as mathscore
CASE WHEN class = 'english' THEN score END as englishscore
FROM
grades
And then bear in mind that MAX() will not consider nulls when doing its work, so it causes the rows to collapse to only those containing values
As an aside - see how i've phrased my answer in terms of giving a sample data and a desired output? Do please write your next SQL question more like that, including the SQL you tried
Edit:
OK, so just apply the pattern in my SQL, to your names. For me, the "name" was the fixed column, the "class" was the one being pivoted out into multiple columns. The general pattern is:
SELECT
fixedColumn1,
fixedColumn2
..
fixedColumnN,
MAX(CASE WHEN column_with_values_that_must_be_headers = a1_value_you_want_in_its_own_column THEN column_with_value_to_appear_as_data1) as column_header_name1,
MAX(CASE WHEN column_with_values_that_must_be_headers = a2_value_you_want_in_its_own_column THEN column_with_value_to_appear_as_data2) as column_header_name2,
...
MAX(CASE WHEN column_with_values_that_must_be_headers = aN_value_you_want_in_its_own_column THEN column_with_value_to_appear_as_dataN) as column_header_nameN
FROM
table
GROUP BY
fixedColumn1,
fixedColumn2
..
fixedColumnN,
Here's a partial; have a go at finishing it off - you'll learn way more about this technique put doing a bit of it, than just hanging around waiting for someone to give you the answer to paste in :)
SELECT
TopologyType,
TopologyName,
MAX(CASE WHEN name = 'client1' THEN name END) AS client1,
... --put more CASE WHENs here
FROM
table
GROUP BY
TopologyType,
TopologyName
The only slight difference between your query and my earlier examples is that I had one column that I wanted to be the header (the name of the class; english / maths) but another column to be the data (the test score of the person in that class) whereas you effectively want the name as the column header AND the data. It's ok though - just follow the pattern above, using the name column as the header and the data:
MAX(CASE WHEN name = 'client1' THEN name END) AS client1,
^^^^^^^^^^^^^^^^ ^^^^
column_header cell_data