oracle中wm_concat()函数

select z.admission_code,z.print_date,z.delivery_code,z.supplier_code,z.supplier_text,
max(z.pxh),max(z.platform_action_text),max(z.confirm_point_text)
from 
(
select t.admission_code,t.print_date,t.delivery_code,t.supplier_code,t.supplier_text,
to_char(WM_CONCAT(to_char(t.PXH)) over (partition by t.admission_code  order by t.pxh))  pxh,
to_char(WM_CONCAT(to_char(t.platform_action_text)) over (order by t.pxh) )  platform_action_text,
to_char(WM_CONCAT(to_char(t.confirm_point_text)) over (order by t.pxh)) confirm_point_text
from 
 (
select a.PXH,m.admission_code,sysdate print_date,m.delivery_code,m.supplier_code,
b.platform_action_text,c.confirm_point_text,d.supplier_text
FROM wm_platform_delivery m 
left join  wm_platform_confirm_info a  on m.admission_code=a.admission_code
left join  WM_PLATFORM_ACTION b on  a.warehouse_code=b.warehouse_code and a.platform_action=b.platform_action    
left join  WM_PLATFORM_CONFIRMATION c on  a.warehouse_code=c.warehouse_code and a.confirm_point=c.confirm_point
left join  wm_supplier d on m.supplier_code=d.supplier_code
WHERE m.ADMISSION_CODE=631  
  ) t
 --上面用了partition by，这里就可以省略 group by t.admission_code,t.print_date,t.delivery_code,t.supplier_code,t.supplier_text
) z
 group by z.admission_code,z.print_date,z.delivery_code,z.supplier_code,z.supplier_text

sql解析：

to_char(WM_CONCAT(to_char(t.PXH)) over (partition by t.admission_code  order by t.pxh))  pxh,
to_char(WM_CONCAT(to_char(t.platform_action_text)) over (order by t.pxh) )  platform_action_text,
to_char(WM_CONCAT(to_char(t.confirm_point_text)) over (order by t.pxh)) confirm_point_text

to_char():加了to_char()，是为了预防字符集原因，合并后的列查出来空的问题

partition by :分组的意思， 等同于group by ，加了这个就可以省略group by

over(): 如果合并后需要排序，就需要加这个反之不需要

原始数据

最终数据（合并并排序）