WITH final AS (
SELECT 'a' AS gname, 'b' AS pname, 'c' AS name, 1 AS gid, 2 AS pid, 3 AS did, 3 AS inp
UNION
SELECT 'a1' AS gname, 'b1' AS pname, 'c1' AS name, 11 AS gid, 21 AS pid, 31 AS did, 3 AS inp
)
Now I want to write a query depending on the inp column value
SELECT
(CASE final.inp
WHEN 1
THEN
(SELECT DISTINCT(final.gid) AS id, final.gname as name FROM
final ORDER BY final.gname)
WHEN 2
THEN
(SELECT DISTINCT(final.pid) AS id, final.pname as name FROM
final WHERE final.gid = $1 ORDER BY final.pname)
WHEN 3
THEN
(SELECT DISTINCT(final.did) AS id, final.name as name FROM
final WHERE final.pid = $1 ORDER BY final.name)
END)
FROM (%s) final;
I know this is not a correct way of writing it.
But this will give you an idea what I am trying to do.
if inp = 1
-> get unique gid's and gname and order by gname
if inp = 2
-> get unique pid's and pname where gid = and order by pname.
I want to write this logic in psql.
I know case When
can have only one column so I have to create two cases for a single final.inp=1
for name and id.
I ain't able to write this logic because of this part
(SELECT DISTINCT(final.pid) AS id, final.pname as name FROM
final WHERE final.gid = $1 ORDER BY final.pname)
WHERE ORDER BY
<--- and passing the same final value to it.
I am not able to implement it. Can anyone help?
CodePudding user response:
Only assumption
select distinct
case when inp = 1 then gid
when inp = 2 then pid
when inp = 3 then did
end new_id,
case when inp = 1 then gname
when inp = 2 then pname
when inp = 3 then name
end as name
from final
order by 2