SQL

Question

我正在尝试从 stackexchange 数据库中获取数据。我的查询是：

select distinct top 50 U.Id, U.DisplayName, U.Reputation,
       Tags = stuff( (SELECT ','+p2.Tags 
                       FROM posts p2 join votes V on p2.id = V.PostId
                       where V.VoteTypeId=5 and V.UserId = U.id
                       order by p2.CreationDate DESC
                       FOR XML PATH, TYPE).value('.[1]','nvarchar(max)')
                    ,1,1,'')
from Users U 
order by U.Reputation DESC;

但是，当我运行 data.stackexchange.com 上的查询时，它显示一条错误消息：Execution Timeout Expired. 有什么方法可以修改查询以优化执行时间，所以我可以运行这个查询成功吗？

Answer 1

您需要排除具有单个 XML 的子查询。这将需要使用结构 <user ...><tag1><tag2>...</user><user>...（使用 EXPLICIT MODE）构建公共 XML。并将其解析为单个用户的行。

with U as(
    select top 50 id, Reputation, DisplayName
      from users order by Reputation DESC
),
V as(
 select U.*,p2.Tags, p2.CreationDate
  from U, posts p2, votes V
 where p2.id = V.PostId and V.VoteTypeId=5 and V.UserId = U.id
),
Q(XML) as(
    select tag,parent,id [user!1!id],Reputation [user!1!Reputation],Name [user!1!Name], T [T!2!!element]
      from (
        select 1 tag, NULL parent, U.id, U.Reputation, U.DisplayName Name, NULL t, NULL dt
          from U
         union all
        select 2, 1, V.id, NULL, NULL, ','+Tags, CreationDate
          from V
     ) X
     order by id, tag, dt desc
       for xml explicit, type
)
select node.value('(@id)[1]','int'),
       node.value('(@Reputation)[1]','int'),
       node.value('(@Name)[1]','nvarchar(max)'),
       stuff(node.value('.[1]','nvarchar(max)'),1,1,'')
  from Q
 cross apply XML.nodes('/user') as Y(node)
 order by 2 desc

SQL - 优化执行时间

SQL - Optimize the execution time

sql-server

data.stackexchange.com