SQL Server 2016：用于XML和具有多连接查询的内容

出于报告目的，我提出了一个查询，列出了尚未提交季度报告的雇主以及尚未提交季度报告的雇主。这是我的疑问：

with employerIds as (select distinct Employerid
                from employerTransaction),
quarters as (select distinct QId
                    from employerTransaction)
select ei.EmployerId, e.EmployerName, q.QId
from employerIds as ei
cross join quarters as q
left join employerTransaction as et
        on et.EmployerId = ei.EmployerId
        and et.QId = q.QId
join employer as e
        on e.EmployerId = ei.EmployerId
where et.employerid is null
group by ei.employerid, q.QId
order by ei.EmployerId, q.QId

以下是查询的结果：

    EmployerId   EmployerName                 QId
    1            Potato Inc                   20193
    1            Potato Inc                   20202
    1            Potato Inc                   20203
    2            Donuts LLC                   20202
    2            Donuts LLC                   20203
    3            Pineapple Logistics          20191
    3            Pineapple Logistics          20192
    3            Pineapple Logistics          20193
    3            Pineapple Logistics          20194

我想为每个EmployerId将QId列合并为一行，如下所示：

     EmployerId   EmployerName                 QId
     1            Potato Inc                   20193, 20202, 20203
     2            Donuts LLC                   20202, 20203
     3            Pineapple Logistics          20191, 20192, 20193, 20194

我使用的是Sql Server 2016，因此不幸的是，我无法利用

string\u agg（）

，而必须将

用于XML

和

stuff（）

。我不知道如何在具有多个联接的查询上使用

stuff（）

。感谢您的帮助。

尝试此解决方案

Select  ei.EmployerId, 
        e.EmployerName ,
        ISNULL(STUFF((Select Distinct ', ' + Cast(QId  As Varchar(500))
                        From employerTransaction As Q
                        Where  et.QId = q.QId
                        FOR XML PATH('')), 1, 1, ''),'-') As [NewQID]
From employerIds As ei
Left Join employerTransaction as et on et.EmployerId = ei.EmployerId and et.QId = q.QId
Join employer as e on e.EmployerId = ei.EmployerId
Where et.employerid is null
Group By ei.employerid, e.EmployerName

---

这是执行该技巧的xml路径

子句<代码>素材

只需去掉前面的“，”

with sampledata as (
   -- your original query here
)
select EmployerId, EmployerName,
   stuff( (select ','+ cast(t2.Qid as varchar(5))
          from sampledata t2
          where t2.EmployerId = t1.EmployerId and t2.EmployerName = t1.EmployerName
          for xml path ('')), 1, 1,'') qids
from sampledata t1
group by EmployerId, EmployerName

返回

EmployerId  EmployerName    qids
1   Potato Inc  20193,20202,20203
2   Donuts LLC  20202,20203
3   Pineapple Logistics 20191,20192,20193,20194

---

您不需要为此使用STUFF函数。您可以使用FOR XML（子）查询来实现这一点。STUFF函数仅用于通过删除前导（或尾随）分隔符来美化最终结果。如果使用逗号和空格（

，“

）作为分隔符，分隔符长度为2个字符。您还应该在STUFF函数的第三个参数中反映这一点。应该是2而不是1。否则，

[NewQID]

列中的结果值将以空格开头。感谢您的回复。Qid正在聚合，但我在运行查询时看到重复的Employerid返回。例如，在我在问题中提供的第二个示例输出中，土豆公司返回三次，甜甜圈有限责任公司返回两次，菠萝物流公司返回四次等等。对于xml路径（“”））.value（'text（）'，'nvarchar（max）'），1，1…@BGonzaga。。完全使用您的示例数据。我没有看到重复的。