Question

My sql query so far has a few joins on tables and the final output looks like this:

FLAG        id      name
----        ---     ----
OK          21      ken
OK          34      mon
OK          51      jil
OK          51      jil
OK          71      jil
OK          80      ron
OK          91      ron

Now I want the FLAGs of duplicate names be shown as ‘dup’ of the lowest id:

FLAG        id      name
----        ---     ----
OK          21      ken
OK          34      mon
OK          51      jil
dup_51      51      jil
dup_51      71      jil
OK          80      ron
dup_80      91      ron

I can do it by using shell/perl script on the records stored in a file, but need to know if it can be possible by manipulating my SQL query.. Thanks for your time and help.

Answer 1

I would suggest a subquery that tracks duplicate entries on name and returns the minimum id for each duplicate. I’ll asume your table is named tbl:

select name, min(id) as dupId
from tbl
group by name
having count(name) > 1

Now you can join this with your original table:

select 
    tbl.*, if(not isnull(dupId), 'Ok', concat('dup_', dupId)) as flag2
from 
    tbl
    left join (
        select name, min(id) as dupId
        from tbl
        group by name
        having count(name) > 1) as a on tbl.name = a.name

This is a way to do it. Hope it helps you.