问题描述
我创建了一个查询来过滤数据。我只需要在最终语句中选择不同的 inv_mast_uid。如果相同的 inv_mast_uid 重复,我只需要选择 1 行。请推荐
with s1
as (select count(supplier_part_no)as count_of_duplicates,supplier_part_no,supplier_id,inv.item_desc,extended_desc
from p21_view_inventory_supplier supp
join p21_view_inv_mast inv
on supp.item_id = inv.item_id
where product_type = 'R'
and supp.delete_flag = 'N'
and inv.delete_flag = 'N'
and other_charge_item = 'N'
group by supplier_part_no,item_desc,extended_desc
having count(supplier_part_no) > 1),s2 as (select supp.inv_mast_uid,inv.item_id,extended_desc
from p21_view_inventory_supplier supp
left join p21_view_inv_mast inv
on supp.item_id = inv.item_id)
select s2.inv_mast_uid,s2.item_id,count_of_duplicates,s1.supplier_part_no,s1.supplier_id,s1.extended_desc,s1.item_desc
from s2
join s1
on s1.supplier_part_no = s2.supplier_part_no
where s1.supplier_id = s2.supplier_id
and s1.item_desc = s2.item_desc
and ( s1.extended_desc = s2.extended_desc
or s1.extended_desc is null
and s2.extended_desc is null )
结果是
预期结果:
希望只有第 1 行和第 2 行。不影响其他 group by 函数。
解决方法
只需使用不同的。它将根据您选择的列返回不同的行。也就是说,重复的行将被删除。
select distinct s2.inv_mast_uid as inv_mast_uid,s2.item_id,count_of_duplicates,s1.supplier_part_no,s1.supplier_id as supplier_id,s1.extended_desc,s1.item_desc
from s2
join s1
on s1.supplier_part_no = s2.supplier_part_no
where s1.supplier_id = s2.supplier_id
and s1.item_desc = s2.item_desc
and ( s1.extended_desc = s2.extended_desc
or (s1.extended_desc is null
and s2.extended_desc is null) )
我猜你在最后一行放错了括号,所以在 or
条件中添加了它。
但是,如果您只查找列 inv_mast_uid
的不同值并且不关心其他列上的值,您可以这样做:
把最后一个select写成s3
,然后把select写成如下:
select p.*
from s3 as p
inner join (select inv_mast_uid,min(supplier_id) from
s3 group by s2.inv_mast_uid) as q
on p.inv_mast_uid= q.inv_mast_uid
and p.supplier_id = q.supplier_id