Group_concat into group_concat with two different separators

I use the follown as one movie's page

Code:

     SELECT
    `mt_films`.`Title1` AS `Title1`,
    GROUP_CONCAT(DISTINCT `t_idiotita`.`idiotita`,
        ': ',
        `t_people`.`person`
        ORDER BY `t_idiotita`.`cnt_gr_id` ASC
        SEPARATOR ' ') AS `kati`
FROM
    ((`t_people`
    JOIN (`st_peoplefilms`
    JOIN `mt_films` ON ((`mt_films`.`ID_films` = `st_peoplefilms`.`ID_films`))) ON ((`t_people`.`ID_person` = `st_peoplefilms`.`ID_person`)))
    JOIN `t_idiotita` ON ((`st_peoplefilms`.`ID_idiotita` = `t_idiotita`.`ID_idiotita`)))
WHERE
    ((`st_peoplefilms`.`ID_idiotita` = 29)
        OR (`st_peoplefilms`.`ID_idiotita` = 33)
        OR (`st_peoplefilms`.`ID_idiotita` = 27)
        OR (`st_peoplefilms`.`ID_idiotita` = 12)
        OR (`st_peoplefilms`.`ID_idiotita` = 43)
        OR (`st_peoplefilms`.`ID_idiotita` = 11)
        OR (`st_peoplefilms`.`ID_idiotita` = 32)
        OR (`st_peoplefilms`.`ID_idiotita` = 5)
        OR (`st_peoplefilms`.`ID_idiotita` = 24)
        OR (`st_peoplefilms`.`ID_idiotita` = 14))
GROUP BY `mt_films`.`Title1`

That returns e.g.

Title1 | Dir.: John | Script: Peter | Script: Anna | ...
Title2 | Dir.: George | Dir.: Paul | Script: Gus | ...

But while for many many movies work more than one person for the same specialty i need a second group_concat to group all the persons for the same specilty. I also need two different separations. The first ": " separates the specialties and the second ", " the persons

Expected result

Title1 | Dir.: John | Script: Peter, Anna | ...
Title2 | Dir.: George, Paul | Script: Gus | ...

Could you pls help?
Thank's

 

Give you an example here:

Code:

select  a,group_concat(concat(b,' ',c)) from
(select emp_dept as a,emp_grade as b,group_concat(distinct emp_sex SEPARATOR ' ') as c from emp_small where emp_dept<'B' group by emp_dept,emp_grade) dt
group by a ;

subquery result:
'AAA', '1000', 'F M'
'AAA', '1001', 'M'
'AAA', '1003', 'M'
'AAA', '1005', 'M'
'AAA', '1006', 'F'
'AAA', '1012', 'M'
'AAA', '1014', 'F M'
'AAA', '1020', 'F'
'AAA', '1025', 'F'
'AAA', '1027', 'F'
'AAA', '1029', 'M'

Final result:
'AAA', '1000 F M,1001 M,1003 M,1005 M,1006 F,1012 M,1014 F M,1020 F,1025 F,1027 F,1029 M
'ABC', '1001 F,1002 F,1011 M,1012 F,1013 M,1027 M,1029 M,1033 M,1040 F,1049 M,1050 F,1052 M
....

 

Hey Richard,
didn't manage to make it works

 

I try my best to understand your SQL, you may try this, the key factor is that whether the grouping of `t_idiotita`.`cnt_gr_id` is applicable, it is up to your judgement.

SELECT
dt.`Title1` AS `Title1`,
GROUP_CONCAT(DISTINCT dt.`idiotita`,
': ',
dt.`person`
ORDER BY dt.`cnt_gr_id` ASC
SEPARATOR ' ') AS `kati` from
(SELECT
`mt_films`.`Title1` AS `Title1`,
`t_idiotita`.`idiotita` as idiotita,
`t_idiotita`.`cnt_gr_id` as `cnt_gr_id`
GROUP_CONCAT(DISTINCT`t_people`.`person`
SEPARATOR ',') AS `persons`
FROM ...
.......
group by `mt_films`.`Title1`,`t_idiotita`.`idiotita`,`t_idiotita`.`cnt_gr_id`) dt
GROUP BY dt.`Title1`